OpenCloudOS-Kernel/security/selinux/hooks.c

5866 lines
141 KiB
C
Raw Normal View History

/*
* NSA Security-Enhanced Linux (SELinux) security module
*
* This file contains the SELinux hook function implementations.
*
* Authors: Stephen Smalley, <sds@epoch.ncsc.mil>
* Chris Vance, <cvance@nai.com>
* Wayne Salamon, <wsalamon@nai.com>
* James Morris <jmorris@redhat.com>
*
* Copyright (C) 2001,2002 Networks Associates Technology, Inc.
* Copyright (C) 2003-2008 Red Hat, Inc., James Morris <jmorris@redhat.com>
* Eric Paris <eparis@redhat.com>
* Copyright (C) 2004-2005 Trusted Computer Solutions, Inc.
* <dgoeddel@trustedcs.com>
* Copyright (C) 2006, 2007 Hewlett-Packard Development Company, L.P.
* Paul Moore <paul.moore@hp.com>
* Copyright (C) 2007 Hitachi Software Engineering Co., Ltd.
* Yuichi Nakamura <ynakam@hitachisoft.jp>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2,
* as published by the Free Software Foundation.
*/
#include <linux/init.h>
#include <linux/kernel.h>
#include <linux/tracehook.h>
#include <linux/errno.h>
#include <linux/sched.h>
#include <linux/security.h>
#include <linux/xattr.h>
#include <linux/capability.h>
#include <linux/unistd.h>
#include <linux/mm.h>
#include <linux/mman.h>
#include <linux/slab.h>
#include <linux/pagemap.h>
#include <linux/swap.h>
#include <linux/spinlock.h>
#include <linux/syscalls.h>
#include <linux/file.h>
#include <linux/fdtable.h>
#include <linux/namei.h>
#include <linux/mount.h>
#include <linux/proc_fs.h>
#include <linux/netfilter_ipv4.h>
#include <linux/netfilter_ipv6.h>
#include <linux/tty.h>
#include <net/icmp.h>
#include <net/ip.h> /* for local_port_range[] */
#include <net/tcp.h> /* struct or_callable used in sock_rcv_skb */
#include <net/net_namespace.h>
#include <net/netlabel.h>
#include <linux/uaccess.h>
#include <asm/ioctls.h>
#include <asm/atomic.h>
#include <linux/bitops.h>
#include <linux/interrupt.h>
#include <linux/netdevice.h> /* for network interface checks */
#include <linux/netlink.h>
#include <linux/tcp.h>
#include <linux/udp.h>
#include <linux/dccp.h>
#include <linux/quota.h>
#include <linux/un.h> /* for Unix socket types */
#include <net/af_unix.h> /* for Unix socket types */
#include <linux/parser.h>
#include <linux/nfs_mount.h>
#include <net/ipv6.h>
#include <linux/hugetlb.h>
#include <linux/personality.h>
#include <linux/sysctl.h>
#include <linux/audit.h>
#include <linux/string.h>
[AF_UNIX]: Datagram getpeersec This patch implements an API whereby an application can determine the label of its peer's Unix datagram sockets via the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of the peer of a Unix datagram socket. The application can then use this security context to determine the security context for processing on behalf of the peer who sent the packet. Patch design and implementation: The design and implementation is very similar to the UDP case for INET sockets. Basically we build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). To retrieve the security context, the application first indicates to the kernel such desire by setting the SO_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for Unix datagram socket should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_SOCKET, SO_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_SOCKET && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } sock_setsockopt is enhanced with a new socket option SOCK_PASSSEC to allow a server socket to receive security context of the peer. Testing: We have tested the patch by setting up Unix datagram client and server applications. We verified that the server can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: Acked-by: James Morris <jmorris@namei.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-30 03:27:47 +08:00
#include <linux/selinux.h>
#include <linux/mutex.h>
timers: fix itimer/many thread hang Overview This patch reworks the handling of POSIX CPU timers, including the ITIMER_PROF, ITIMER_VIRT timers and rlimit handling. It was put together with the help of Roland McGrath, the owner and original writer of this code. The problem we ran into, and the reason for this rework, has to do with using a profiling timer in a process with a large number of threads. It appears that the performance of the old implementation of run_posix_cpu_timers() was at least O(n*3) (where "n" is the number of threads in a process) or worse. Everything is fine with an increasing number of threads until the time taken for that routine to run becomes the same as or greater than the tick time, at which point things degrade rather quickly. This patch fixes bug 9906, "Weird hang with NPTL and SIGPROF." Code Changes This rework corrects the implementation of run_posix_cpu_timers() to make it run in constant time for a particular machine. (Performance may vary between one machine and another depending upon whether the kernel is built as single- or multiprocessor and, in the latter case, depending upon the number of running processors.) To do this, at each tick we now update fields in signal_struct as well as task_struct. The run_posix_cpu_timers() function uses those fields to make its decisions. We define a new structure, "task_cputime," to contain user, system and scheduler times and use these in appropriate places: struct task_cputime { cputime_t utime; cputime_t stime; unsigned long long sum_exec_runtime; }; This is included in the structure "thread_group_cputime," which is a new substructure of signal_struct and which varies for uniprocessor versus multiprocessor kernels. For uniprocessor kernels, it uses "task_cputime" as a simple substructure, while for multiprocessor kernels it is a pointer: struct thread_group_cputime { struct task_cputime totals; }; struct thread_group_cputime { struct task_cputime *totals; }; We also add a new task_cputime substructure directly to signal_struct, to cache the earliest expiration of process-wide timers, and task_cputime also replaces the it_*_expires fields of task_struct (used for earliest expiration of thread timers). The "thread_group_cputime" structure contains process-wide timers that are updated via account_user_time() and friends. In the non-SMP case the structure is a simple aggregator; unfortunately in the SMP case that simplicity was not achievable due to cache-line contention between CPUs (in one measured case performance was actually _worse_ on a 16-cpu system than the same test on a 4-cpu system, due to this contention). For SMP, the thread_group_cputime counters are maintained as a per-cpu structure allocated using alloc_percpu(). The timer functions update only the timer field in the structure corresponding to the running CPU, obtained using per_cpu_ptr(). We define a set of inline functions in sched.h that we use to maintain the thread_group_cputime structure and hide the differences between UP and SMP implementations from the rest of the kernel. The thread_group_cputime_init() function initializes the thread_group_cputime structure for the given task. The thread_group_cputime_alloc() is a no-op for UP; for SMP it calls the out-of-line function thread_group_cputime_alloc_smp() to allocate and fill in the per-cpu structures and fields. The thread_group_cputime_free() function, also a no-op for UP, in SMP frees the per-cpu structures. The thread_group_cputime_clone_thread() function (also a UP no-op) for SMP calls thread_group_cputime_alloc() if the per-cpu structures haven't yet been allocated. The thread_group_cputime() function fills the task_cputime structure it is passed with the contents of the thread_group_cputime fields; in UP it's that simple but in SMP it must also safely check that tsk->signal is non-NULL (if it is it just uses the appropriate fields of task_struct) and, if so, sums the per-cpu values for each online CPU. Finally, the three functions account_group_user_time(), account_group_system_time() and account_group_exec_runtime() are used by timer functions to update the respective fields of the thread_group_cputime structure. Non-SMP operation is trivial and will not be mentioned further. The per-cpu structure is always allocated when a task creates its first new thread, via a call to thread_group_cputime_clone_thread() from copy_signal(). It is freed at process exit via a call to thread_group_cputime_free() from cleanup_signal(). All functions that formerly summed utime/stime/sum_sched_runtime values from from all threads in the thread group now use thread_group_cputime() to snapshot the values in the thread_group_cputime structure or the values in the task structure itself if the per-cpu structure hasn't been allocated. Finally, the code in kernel/posix-cpu-timers.c has changed quite a bit. The run_posix_cpu_timers() function has been split into a fast path and a slow path; the former safely checks whether there are any expired thread timers and, if not, just returns, while the slow path does the heavy lifting. With the dedicated thread group fields, timers are no longer "rebalanced" and the process_timer_rebalance() function and related code has gone away. All summing loops are gone and all code that used them now uses the thread_group_cputime() inline. When process-wide timers are set, the new task_cputime structure in signal_struct is used to cache the earliest expiration; this is checked in the fast path. Performance The fix appears not to add significant overhead to existing operations. It generally performs the same as the current code except in two cases, one in which it performs slightly worse (Case 5 below) and one in which it performs very significantly better (Case 2 below). Overall it's a wash except in those two cases. I've since done somewhat more involved testing on a dual-core Opteron system. Case 1: With no itimer running, for a test with 100,000 threads, the fixed kernel took 1428.5 seconds, 513 seconds more than the unfixed system, all of which was spent in the system. There were twice as many voluntary context switches with the fix as without it. Case 2: With an itimer running at .01 second ticks and 4000 threads (the most an unmodified kernel can handle), the fixed kernel ran the test in eight percent of the time (5.8 seconds as opposed to 70 seconds) and had better tick accuracy (.012 seconds per tick as opposed to .023 seconds per tick). Case 3: A 4000-thread test with an initial timer tick of .01 second and an interval of 10,000 seconds (i.e. a timer that ticks only once) had very nearly the same performance in both cases: 6.3 seconds elapsed for the fixed kernel versus 5.5 seconds for the unfixed kernel. With fewer threads (eight in these tests), the Case 1 test ran in essentially the same time on both the modified and unmodified kernels (5.2 seconds versus 5.8 seconds). The Case 2 test ran in about the same time as well, 5.9 seconds versus 5.4 seconds but again with much better tick accuracy, .013 seconds per tick versus .025 seconds per tick for the unmodified kernel. Since the fix affected the rlimit code, I also tested soft and hard CPU limits. Case 4: With a hard CPU limit of 20 seconds and eight threads (and an itimer running), the modified kernel was very slightly favored in that while it killed the process in 19.997 seconds of CPU time (5.002 seconds of wall time), only .003 seconds of that was system time, the rest was user time. The unmodified kernel killed the process in 20.001 seconds of CPU (5.014 seconds of wall time) of which .016 seconds was system time. Really, though, the results were too close to call. The results were essentially the same with no itimer running. Case 5: With a soft limit of 20 seconds and a hard limit of 2000 seconds (where the hard limit would never be reached) and an itimer running, the modified kernel exhibited worse tick accuracy than the unmodified kernel: .050 seconds/tick versus .028 seconds/tick. Otherwise, performance was almost indistinguishable. With no itimer running this test exhibited virtually identical behavior and times in both cases. In times past I did some limited performance testing. those results are below. On a four-cpu Opteron system without this fix, a sixteen-thread test executed in 3569.991 seconds, of which user was 3568.435s and system was 1.556s. On the same system with the fix, user and elapsed time were about the same, but system time dropped to 0.007 seconds. Performance with eight, four and one thread were comparable. Interestingly, the timer ticks with the fix seemed more accurate: The sixteen-thread test with the fix received 149543 ticks for 0.024 seconds per tick, while the same test without the fix received 58720 for 0.061 seconds per tick. Both cases were configured for an interval of 0.01 seconds. Again, the other tests were comparable. Each thread in this test computed the primes up to 25,000,000. I also did a test with a large number of threads, 100,000 threads, which is impossible without the fix. In this case each thread computed the primes only up to 10,000 (to make the runtime manageable). System time dominated, at 1546.968 seconds out of a total 2176.906 seconds (giving a user time of 629.938s). It received 147651 ticks for 0.015 seconds per tick, still quite accurate. There is obviously no comparable test without the fix. Signed-off-by: Frank Mayhar <fmayhar@google.com> Cc: Roland McGrath <roland@redhat.com> Cc: Alexey Dobriyan <adobriyan@gmail.com> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-09-13 00:54:39 +08:00
#include <linux/posix-timers.h>
#include "avc.h"
#include "objsec.h"
#include "netif.h"
#include "netnode.h"
#include "netport.h"
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
#include "xfrm.h"
#include "netlabel.h"
#include "audit.h"
#define XATTR_SELINUX_SUFFIX "selinux"
#define XATTR_NAME_SELINUX XATTR_SECURITY_PREFIX XATTR_SELINUX_SUFFIX
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
#define NUM_SEL_MNT_OPTS 4
extern unsigned int policydb_loaded_version;
extern int selinux_nlmsg_lookup(u16 sclass, u16 nlmsg_type, u32 *perm);
extern int selinux_compat_net;
extern struct security_operations *security_ops;
/* SECMARK reference count */
atomic_t selinux_secmark_refcount = ATOMIC_INIT(0);
#ifdef CONFIG_SECURITY_SELINUX_DEVELOP
int selinux_enforcing;
static int __init enforcing_setup(char *str)
{
unsigned long enforcing;
if (!strict_strtoul(str, 0, &enforcing))
selinux_enforcing = enforcing ? 1 : 0;
return 1;
}
__setup("enforcing=", enforcing_setup);
#endif
#ifdef CONFIG_SECURITY_SELINUX_BOOTPARAM
int selinux_enabled = CONFIG_SECURITY_SELINUX_BOOTPARAM_VALUE;
static int __init selinux_enabled_setup(char *str)
{
unsigned long enabled;
if (!strict_strtoul(str, 0, &enabled))
selinux_enabled = enabled ? 1 : 0;
return 1;
}
__setup("selinux=", selinux_enabled_setup);
#else
int selinux_enabled = 1;
#endif
/*
* Minimal support for a secondary security module,
* just to allow the use of the capability module.
*/
static struct security_operations *secondary_ops;
/* Lists of inode and superblock security structures initialized
before the policy was loaded. */
static LIST_HEAD(superblock_security_head);
static DEFINE_SPINLOCK(sb_security_lock);
static struct kmem_cache *sel_inode_cache;
/**
* selinux_secmark_enabled - Check to see if SECMARK is currently enabled
*
* Description:
* This function checks the SECMARK reference counter to see if any SECMARK
* targets are currently configured, if the reference counter is greater than
* zero SECMARK is considered to be enabled. Returns true (1) if SECMARK is
* enabled, false (0) if SECMARK is disabled.
*
*/
static int selinux_secmark_enabled(void)
{
return (atomic_read(&selinux_secmark_refcount) > 0);
}
/* Allocate and free functions for each kind of security blob. */
static int task_alloc_security(struct task_struct *task)
{
struct task_security_struct *tsec;
tsec = kzalloc(sizeof(struct task_security_struct), GFP_KERNEL);
if (!tsec)
return -ENOMEM;
tsec->osid = tsec->sid = SECINITSID_UNLABELED;
task->security = tsec;
return 0;
}
static void task_free_security(struct task_struct *task)
{
struct task_security_struct *tsec = task->security;
task->security = NULL;
kfree(tsec);
}
static int inode_alloc_security(struct inode *inode)
{
struct task_security_struct *tsec = current->security;
struct inode_security_struct *isec;
isec = kmem_cache_zalloc(sel_inode_cache, GFP_NOFS);
if (!isec)
return -ENOMEM;
mutex_init(&isec->lock);
INIT_LIST_HEAD(&isec->list);
isec->inode = inode;
isec->sid = SECINITSID_UNLABELED;
isec->sclass = SECCLASS_FILE;
isec->task_sid = tsec->sid;
inode->i_security = isec;
return 0;
}
static void inode_free_security(struct inode *inode)
{
struct inode_security_struct *isec = inode->i_security;
struct superblock_security_struct *sbsec = inode->i_sb->s_security;
spin_lock(&sbsec->isec_lock);
if (!list_empty(&isec->list))
list_del_init(&isec->list);
spin_unlock(&sbsec->isec_lock);
inode->i_security = NULL;
kmem_cache_free(sel_inode_cache, isec);
}
static int file_alloc_security(struct file *file)
{
struct task_security_struct *tsec = current->security;
struct file_security_struct *fsec;
fsec = kzalloc(sizeof(struct file_security_struct), GFP_KERNEL);
if (!fsec)
return -ENOMEM;
fsec->sid = tsec->sid;
fsec->fown_sid = tsec->sid;
file->f_security = fsec;
return 0;
}
static void file_free_security(struct file *file)
{
struct file_security_struct *fsec = file->f_security;
file->f_security = NULL;
kfree(fsec);
}
static int superblock_alloc_security(struct super_block *sb)
{
struct superblock_security_struct *sbsec;
sbsec = kzalloc(sizeof(struct superblock_security_struct), GFP_KERNEL);
if (!sbsec)
return -ENOMEM;
mutex_init(&sbsec->lock);
INIT_LIST_HEAD(&sbsec->list);
INIT_LIST_HEAD(&sbsec->isec_head);
spin_lock_init(&sbsec->isec_lock);
sbsec->sb = sb;
sbsec->sid = SECINITSID_UNLABELED;
sbsec->def_sid = SECINITSID_FILE;
sbsec->mntpoint_sid = SECINITSID_UNLABELED;
sb->s_security = sbsec;
return 0;
}
static void superblock_free_security(struct super_block *sb)
{
struct superblock_security_struct *sbsec = sb->s_security;
spin_lock(&sb_security_lock);
if (!list_empty(&sbsec->list))
list_del_init(&sbsec->list);
spin_unlock(&sb_security_lock);
sb->s_security = NULL;
kfree(sbsec);
}
static int sk_alloc_security(struct sock *sk, int family, gfp_t priority)
{
struct sk_security_struct *ssec;
ssec = kzalloc(sizeof(*ssec), priority);
if (!ssec)
return -ENOMEM;
ssec->peer_sid = SECINITSID_UNLABELED;
ssec->sid = SECINITSID_UNLABELED;
sk->sk_security = ssec;
selinux_netlbl_sk_security_reset(ssec, family);
return 0;
}
static void sk_free_security(struct sock *sk)
{
struct sk_security_struct *ssec = sk->sk_security;
sk->sk_security = NULL;
selinux_netlbl_sk_security_free(ssec);
kfree(ssec);
}
/* The security server must be initialized before
any labeling or access decisions can be provided. */
extern int ss_initialized;
/* The file system's label must be initialized prior to use. */
static char *labeling_behaviors[6] = {
"uses xattr",
"uses transition SIDs",
"uses task SIDs",
"uses genfs_contexts",
"not configured for labeling",
"uses mountpoint labeling",
};
static int inode_doinit_with_dentry(struct inode *inode, struct dentry *opt_dentry);
static inline int inode_doinit(struct inode *inode)
{
return inode_doinit_with_dentry(inode, NULL);
}
enum {
Opt_error = -1,
Opt_context = 1,
Opt_fscontext = 2,
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
Opt_defcontext = 3,
Opt_rootcontext = 4,
};
static const match_table_t tokens = {
{Opt_context, CONTEXT_STR "%s"},
{Opt_fscontext, FSCONTEXT_STR "%s"},
{Opt_defcontext, DEFCONTEXT_STR "%s"},
{Opt_rootcontext, ROOTCONTEXT_STR "%s"},
{Opt_error, NULL},
};
#define SEL_MOUNT_FAIL_MSG "SELinux: duplicate or incompatible mount options\n"
static int may_context_mount_sb_relabel(u32 sid,
struct superblock_security_struct *sbsec,
struct task_security_struct *tsec)
{
int rc;
rc = avc_has_perm(tsec->sid, sbsec->sid, SECCLASS_FILESYSTEM,
FILESYSTEM__RELABELFROM, NULL);
if (rc)
return rc;
rc = avc_has_perm(tsec->sid, sid, SECCLASS_FILESYSTEM,
FILESYSTEM__RELABELTO, NULL);
return rc;
}
static int may_context_mount_inode_relabel(u32 sid,
struct superblock_security_struct *sbsec,
struct task_security_struct *tsec)
{
int rc;
rc = avc_has_perm(tsec->sid, sbsec->sid, SECCLASS_FILESYSTEM,
FILESYSTEM__RELABELFROM, NULL);
if (rc)
return rc;
rc = avc_has_perm(sid, sbsec->sid, SECCLASS_FILESYSTEM,
FILESYSTEM__ASSOCIATE, NULL);
return rc;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
static int sb_finish_set_opts(struct super_block *sb)
{
struct superblock_security_struct *sbsec = sb->s_security;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
struct dentry *root = sb->s_root;
struct inode *root_inode = root->d_inode;
int rc = 0;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (sbsec->behavior == SECURITY_FS_USE_XATTR) {
/* Make sure that the xattr handler exists and that no
error other than -ENODATA is returned by getxattr on
the root directory. -ENODATA is ok, as this may be
the first boot of the SELinux kernel before we have
assigned xattr values to the filesystem. */
if (!root_inode->i_op->getxattr) {
printk(KERN_WARNING "SELinux: (dev %s, type %s) has no "
"xattr support\n", sb->s_id, sb->s_type->name);
rc = -EOPNOTSUPP;
goto out;
}
rc = root_inode->i_op->getxattr(root, XATTR_NAME_SELINUX, NULL, 0);
if (rc < 0 && rc != -ENODATA) {
if (rc == -EOPNOTSUPP)
printk(KERN_WARNING "SELinux: (dev %s, type "
"%s) has no security xattr handler\n",
sb->s_id, sb->s_type->name);
else
printk(KERN_WARNING "SELinux: (dev %s, type "
"%s) getxattr errno %d\n", sb->s_id,
sb->s_type->name, -rc);
goto out;
}
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
sbsec->initialized = 1;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (sbsec->behavior > ARRAY_SIZE(labeling_behaviors))
printk(KERN_ERR "SELinux: initialized (dev %s, type %s), unknown behavior\n",
sb->s_id, sb->s_type->name);
else
printk(KERN_DEBUG "SELinux: initialized (dev %s, type %s), %s\n",
sb->s_id, sb->s_type->name,
labeling_behaviors[sbsec->behavior-1]);
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
/* Initialize the root inode. */
rc = inode_doinit_with_dentry(root_inode, root);
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
/* Initialize any other inodes associated with the superblock, e.g.
inodes created prior to initial policy load or inodes created
during get_sb by a pseudo filesystem that directly
populates itself. */
spin_lock(&sbsec->isec_lock);
next_inode:
if (!list_empty(&sbsec->isec_head)) {
struct inode_security_struct *isec =
list_entry(sbsec->isec_head.next,
struct inode_security_struct, list);
struct inode *inode = isec->inode;
spin_unlock(&sbsec->isec_lock);
inode = igrab(inode);
if (inode) {
if (!IS_PRIVATE(inode))
inode_doinit(inode);
iput(inode);
}
spin_lock(&sbsec->isec_lock);
list_del_init(&isec->list);
goto next_inode;
}
spin_unlock(&sbsec->isec_lock);
out:
return rc;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
/*
* This function should allow an FS to ask what it's mount security
* options were so it can use those later for submounts, displaying
* mount options, or whatever.
*/
static int selinux_get_mnt_opts(const struct super_block *sb,
struct security_mnt_opts *opts)
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
{
int rc = 0, i;
struct superblock_security_struct *sbsec = sb->s_security;
char *context = NULL;
u32 len;
char tmp;
security_init_mnt_opts(opts);
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (!sbsec->initialized)
return -EINVAL;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (!ss_initialized)
return -EINVAL;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
/*
* if we ever use sbsec flags for anything other than tracking mount
* settings this is going to need a mask
*/
tmp = sbsec->flags;
/* count the number of mount options for this sb */
for (i = 0; i < 8; i++) {
if (tmp & 0x01)
opts->num_mnt_opts++;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
tmp >>= 1;
}
opts->mnt_opts = kcalloc(opts->num_mnt_opts, sizeof(char *), GFP_ATOMIC);
if (!opts->mnt_opts) {
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
rc = -ENOMEM;
goto out_free;
}
opts->mnt_opts_flags = kcalloc(opts->num_mnt_opts, sizeof(int), GFP_ATOMIC);
if (!opts->mnt_opts_flags) {
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
rc = -ENOMEM;
goto out_free;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
i = 0;
if (sbsec->flags & FSCONTEXT_MNT) {
rc = security_sid_to_context(sbsec->sid, &context, &len);
if (rc)
goto out_free;
opts->mnt_opts[i] = context;
opts->mnt_opts_flags[i++] = FSCONTEXT_MNT;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
if (sbsec->flags & CONTEXT_MNT) {
rc = security_sid_to_context(sbsec->mntpoint_sid, &context, &len);
if (rc)
goto out_free;
opts->mnt_opts[i] = context;
opts->mnt_opts_flags[i++] = CONTEXT_MNT;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
if (sbsec->flags & DEFCONTEXT_MNT) {
rc = security_sid_to_context(sbsec->def_sid, &context, &len);
if (rc)
goto out_free;
opts->mnt_opts[i] = context;
opts->mnt_opts_flags[i++] = DEFCONTEXT_MNT;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
if (sbsec->flags & ROOTCONTEXT_MNT) {
struct inode *root = sbsec->sb->s_root->d_inode;
struct inode_security_struct *isec = root->i_security;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
rc = security_sid_to_context(isec->sid, &context, &len);
if (rc)
goto out_free;
opts->mnt_opts[i] = context;
opts->mnt_opts_flags[i++] = ROOTCONTEXT_MNT;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
BUG_ON(i != opts->num_mnt_opts);
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
return 0;
out_free:
security_free_mnt_opts(opts);
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
return rc;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
static int bad_option(struct superblock_security_struct *sbsec, char flag,
u32 old_sid, u32 new_sid)
{
/* check if the old mount command had the same options */
if (sbsec->initialized)
if (!(sbsec->flags & flag) ||
(old_sid != new_sid))
return 1;
/* check if we were passed the same options twice,
* aka someone passed context=a,context=b
*/
if (!sbsec->initialized)
if (sbsec->flags & flag)
return 1;
return 0;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
/*
* Allow filesystems with binary mount data to explicitly set mount point
* labeling information.
*/
static int selinux_set_mnt_opts(struct super_block *sb,
struct security_mnt_opts *opts)
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
{
int rc = 0, i;
struct task_security_struct *tsec = current->security;
struct superblock_security_struct *sbsec = sb->s_security;
const char *name = sb->s_type->name;
struct inode *inode = sbsec->sb->s_root->d_inode;
struct inode_security_struct *root_isec = inode->i_security;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
u32 fscontext_sid = 0, context_sid = 0, rootcontext_sid = 0;
u32 defcontext_sid = 0;
char **mount_options = opts->mnt_opts;
int *flags = opts->mnt_opts_flags;
int num_opts = opts->num_mnt_opts;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
mutex_lock(&sbsec->lock);
if (!ss_initialized) {
if (!num_opts) {
/* Defer initialization until selinux_complete_init,
after the initial policy is loaded and the security
server is ready to handle calls. */
spin_lock(&sb_security_lock);
if (list_empty(&sbsec->list))
list_add(&sbsec->list, &superblock_security_head);
spin_unlock(&sb_security_lock);
goto out;
}
rc = -EINVAL;
printk(KERN_WARNING "SELinux: Unable to set superblock options "
"before the security server is initialized\n");
goto out;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
/*
* Binary mount data FS will come through this function twice. Once
* from an explicit call and once from the generic calls from the vfs.
* Since the generic VFS calls will not contain any security mount data
* we need to skip the double mount verification.
*
* This does open a hole in which we will not notice if the first
* mount using this sb set explict options and a second mount using
* this sb does not set any security options. (The first options
* will be used for both mounts)
*/
if (sbsec->initialized && (sb->s_type->fs_flags & FS_BINARY_MOUNTDATA)
&& (num_opts == 0))
goto out;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
/*
* parse the mount options, check if they are valid sids.
* also check if someone is trying to mount the same sb more
* than once with different security options.
*/
for (i = 0; i < num_opts; i++) {
u32 sid;
rc = security_context_to_sid(mount_options[i],
strlen(mount_options[i]), &sid);
if (rc) {
printk(KERN_WARNING "SELinux: security_context_to_sid"
"(%s) failed for (dev %s, type %s) errno=%d\n",
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
mount_options[i], sb->s_id, name, rc);
goto out;
}
switch (flags[i]) {
case FSCONTEXT_MNT:
fscontext_sid = sid;
if (bad_option(sbsec, FSCONTEXT_MNT, sbsec->sid,
fscontext_sid))
goto out_double_mount;
sbsec->flags |= FSCONTEXT_MNT;
break;
case CONTEXT_MNT:
context_sid = sid;
if (bad_option(sbsec, CONTEXT_MNT, sbsec->mntpoint_sid,
context_sid))
goto out_double_mount;
sbsec->flags |= CONTEXT_MNT;
break;
case ROOTCONTEXT_MNT:
rootcontext_sid = sid;
if (bad_option(sbsec, ROOTCONTEXT_MNT, root_isec->sid,
rootcontext_sid))
goto out_double_mount;
sbsec->flags |= ROOTCONTEXT_MNT;
break;
case DEFCONTEXT_MNT:
defcontext_sid = sid;
if (bad_option(sbsec, DEFCONTEXT_MNT, sbsec->def_sid,
defcontext_sid))
goto out_double_mount;
sbsec->flags |= DEFCONTEXT_MNT;
break;
default:
rc = -EINVAL;
goto out;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
if (sbsec->initialized) {
/* previously mounted with options, but not on this attempt? */
if (sbsec->flags && !num_opts)
goto out_double_mount;
rc = 0;
goto out;
}
if (strcmp(sb->s_type->name, "proc") == 0)
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
sbsec->proc = 1;
/* Determine the labeling behavior to use for this filesystem type. */
rc = security_fs_use(sb->s_type->name, &sbsec->behavior, &sbsec->sid);
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (rc) {
printk(KERN_WARNING "%s: security_fs_use(%s) returned %d\n",
__func__, sb->s_type->name, rc);
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
goto out;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
/* sets the context of the superblock for the fs being mounted. */
if (fscontext_sid) {
rc = may_context_mount_sb_relabel(fscontext_sid, sbsec, tsec);
if (rc)
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
goto out;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
sbsec->sid = fscontext_sid;
}
/*
* Switch to using mount point labeling behavior.
* sets the label used on all file below the mountpoint, and will set
* the superblock context if not already set.
*/
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (context_sid) {
if (!fscontext_sid) {
rc = may_context_mount_sb_relabel(context_sid, sbsec, tsec);
if (rc)
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
goto out;
sbsec->sid = context_sid;
} else {
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
rc = may_context_mount_inode_relabel(context_sid, sbsec, tsec);
if (rc)
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
goto out;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (!rootcontext_sid)
rootcontext_sid = context_sid;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
sbsec->mntpoint_sid = context_sid;
sbsec->behavior = SECURITY_FS_USE_MNTPOINT;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (rootcontext_sid) {
rc = may_context_mount_inode_relabel(rootcontext_sid, sbsec, tsec);
if (rc)
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
goto out;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
root_isec->sid = rootcontext_sid;
root_isec->initialized = 1;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (defcontext_sid) {
if (sbsec->behavior != SECURITY_FS_USE_XATTR) {
rc = -EINVAL;
printk(KERN_WARNING "SELinux: defcontext option is "
"invalid for this filesystem type\n");
goto out;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (defcontext_sid != sbsec->def_sid) {
rc = may_context_mount_inode_relabel(defcontext_sid,
sbsec, tsec);
if (rc)
goto out;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
sbsec->def_sid = defcontext_sid;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
rc = sb_finish_set_opts(sb);
out:
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
mutex_unlock(&sbsec->lock);
return rc;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
out_double_mount:
rc = -EINVAL;
printk(KERN_WARNING "SELinux: mount invalid. Same superblock, different "
"security settings for (dev %s, type %s)\n", sb->s_id, name);
goto out;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
static void selinux_sb_clone_mnt_opts(const struct super_block *oldsb,
struct super_block *newsb)
{
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
const struct superblock_security_struct *oldsbsec = oldsb->s_security;
struct superblock_security_struct *newsbsec = newsb->s_security;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
int set_fscontext = (oldsbsec->flags & FSCONTEXT_MNT);
int set_context = (oldsbsec->flags & CONTEXT_MNT);
int set_rootcontext = (oldsbsec->flags & ROOTCONTEXT_MNT);
/*
* if the parent was able to be mounted it clearly had no special lsm
* mount options. thus we can safely put this sb on the list and deal
* with it later
*/
if (!ss_initialized) {
spin_lock(&sb_security_lock);
if (list_empty(&newsbsec->list))
list_add(&newsbsec->list, &superblock_security_head);
spin_unlock(&sb_security_lock);
return;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
/* how can we clone if the old one wasn't set up?? */
BUG_ON(!oldsbsec->initialized);
/* if fs is reusing a sb, just let its options stand... */
if (newsbsec->initialized)
return;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
mutex_lock(&newsbsec->lock);
newsbsec->flags = oldsbsec->flags;
newsbsec->sid = oldsbsec->sid;
newsbsec->def_sid = oldsbsec->def_sid;
newsbsec->behavior = oldsbsec->behavior;
if (set_context) {
u32 sid = oldsbsec->mntpoint_sid;
if (!set_fscontext)
newsbsec->sid = sid;
if (!set_rootcontext) {
struct inode *newinode = newsb->s_root->d_inode;
struct inode_security_struct *newisec = newinode->i_security;
newisec->sid = sid;
}
newsbsec->mntpoint_sid = sid;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (set_rootcontext) {
const struct inode *oldinode = oldsb->s_root->d_inode;
const struct inode_security_struct *oldisec = oldinode->i_security;
struct inode *newinode = newsb->s_root->d_inode;
struct inode_security_struct *newisec = newinode->i_security;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
newisec->sid = oldisec->sid;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
sb_finish_set_opts(newsb);
mutex_unlock(&newsbsec->lock);
}
static int selinux_parse_opts_str(char *options,
struct security_mnt_opts *opts)
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
{
char *p;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
char *context = NULL, *defcontext = NULL;
char *fscontext = NULL, *rootcontext = NULL;
int rc, num_mnt_opts = 0;
opts->num_mnt_opts = 0;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
/* Standard string-based options. */
while ((p = strsep(&options, "|")) != NULL) {
int token;
substring_t args[MAX_OPT_ARGS];
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (!*p)
continue;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
token = match_token(p, tokens, args);
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
switch (token) {
case Opt_context:
if (context || defcontext) {
rc = -EINVAL;
printk(KERN_WARNING SEL_MOUNT_FAIL_MSG);
goto out_err;
}
context = match_strdup(&args[0]);
if (!context) {
rc = -ENOMEM;
goto out_err;
}
break;
case Opt_fscontext:
if (fscontext) {
rc = -EINVAL;
printk(KERN_WARNING SEL_MOUNT_FAIL_MSG);
goto out_err;
}
fscontext = match_strdup(&args[0]);
if (!fscontext) {
rc = -ENOMEM;
goto out_err;
}
break;
case Opt_rootcontext:
if (rootcontext) {
rc = -EINVAL;
printk(KERN_WARNING SEL_MOUNT_FAIL_MSG);
goto out_err;
}
rootcontext = match_strdup(&args[0]);
if (!rootcontext) {
rc = -ENOMEM;
goto out_err;
}
break;
case Opt_defcontext:
if (context || defcontext) {
rc = -EINVAL;
printk(KERN_WARNING SEL_MOUNT_FAIL_MSG);
goto out_err;
}
defcontext = match_strdup(&args[0]);
if (!defcontext) {
rc = -ENOMEM;
goto out_err;
}
break;
default:
rc = -EINVAL;
printk(KERN_WARNING "SELinux: unknown mount option\n");
goto out_err;
}
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
rc = -ENOMEM;
opts->mnt_opts = kcalloc(NUM_SEL_MNT_OPTS, sizeof(char *), GFP_ATOMIC);
if (!opts->mnt_opts)
goto out_err;
opts->mnt_opts_flags = kcalloc(NUM_SEL_MNT_OPTS, sizeof(int), GFP_ATOMIC);
if (!opts->mnt_opts_flags) {
kfree(opts->mnt_opts);
goto out_err;
}
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
if (fscontext) {
opts->mnt_opts[num_mnt_opts] = fscontext;
opts->mnt_opts_flags[num_mnt_opts++] = FSCONTEXT_MNT;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
if (context) {
opts->mnt_opts[num_mnt_opts] = context;
opts->mnt_opts_flags[num_mnt_opts++] = CONTEXT_MNT;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
if (rootcontext) {
opts->mnt_opts[num_mnt_opts] = rootcontext;
opts->mnt_opts_flags[num_mnt_opts++] = ROOTCONTEXT_MNT;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
if (defcontext) {
opts->mnt_opts[num_mnt_opts] = defcontext;
opts->mnt_opts_flags[num_mnt_opts++] = DEFCONTEXT_MNT;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
}
opts->num_mnt_opts = num_mnt_opts;
return 0;
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
out_err:
kfree(context);
kfree(defcontext);
kfree(fscontext);
kfree(rootcontext);
return rc;
}
/*
* string mount options parsing and call set the sbsec
*/
static int superblock_doinit(struct super_block *sb, void *data)
{
int rc = 0;
char *options = data;
struct security_mnt_opts opts;
security_init_mnt_opts(&opts);
if (!data)
goto out;
BUG_ON(sb->s_type->fs_flags & FS_BINARY_MOUNTDATA);
rc = selinux_parse_opts_str(options, &opts);
if (rc)
goto out_err;
out:
rc = selinux_set_mnt_opts(sb, &opts);
out_err:
security_free_mnt_opts(&opts);
return rc;
}
static void selinux_write_opts(struct seq_file *m,
struct security_mnt_opts *opts)
{
int i;
char *prefix;
for (i = 0; i < opts->num_mnt_opts; i++) {
char *has_comma = strchr(opts->mnt_opts[i], ',');
switch (opts->mnt_opts_flags[i]) {
case CONTEXT_MNT:
prefix = CONTEXT_STR;
break;
case FSCONTEXT_MNT:
prefix = FSCONTEXT_STR;
break;
case ROOTCONTEXT_MNT:
prefix = ROOTCONTEXT_STR;
break;
case DEFCONTEXT_MNT:
prefix = DEFCONTEXT_STR;
break;
default:
BUG();
};
/* we need a comma before each option */
seq_putc(m, ',');
seq_puts(m, prefix);
if (has_comma)
seq_putc(m, '\"');
seq_puts(m, opts->mnt_opts[i]);
if (has_comma)
seq_putc(m, '\"');
}
}
static int selinux_sb_show_options(struct seq_file *m, struct super_block *sb)
{
struct security_mnt_opts opts;
int rc;
rc = selinux_get_mnt_opts(sb, &opts);
if (rc) {
/* before policy load we may get EINVAL, don't show anything */
if (rc == -EINVAL)
rc = 0;
return rc;
}
selinux_write_opts(m, &opts);
security_free_mnt_opts(&opts);
return rc;
}
static inline u16 inode_mode_to_security_class(umode_t mode)
{
switch (mode & S_IFMT) {
case S_IFSOCK:
return SECCLASS_SOCK_FILE;
case S_IFLNK:
return SECCLASS_LNK_FILE;
case S_IFREG:
return SECCLASS_FILE;
case S_IFBLK:
return SECCLASS_BLK_FILE;
case S_IFDIR:
return SECCLASS_DIR;
case S_IFCHR:
return SECCLASS_CHR_FILE;
case S_IFIFO:
return SECCLASS_FIFO_FILE;
}
return SECCLASS_FILE;
}
[PATCH] SELinux - fix SCTP socket bug and general IP protocol handling The following patch updates the way SELinux classifies and handles IP based protocols. Currently, IP sockets are classified by SELinux as being either TCP, UDP or 'Raw', the latter being a default for IP socket that is not TCP or UDP. The classification code is out of date and uses only the socket type parameter to socket(2) to determine the class of IP socket. So, any socket created with SOCK_STREAM will be classified by SELinux as TCP, and SOCK_DGRAM as UDP. Also, other socket types such as SOCK_SEQPACKET and SOCK_DCCP are currently ignored by SELinux, which classifies them as generic sockets, which means they don't even get basic IP level checking. This patch changes the SELinux IP socket classification logic, so that only an IPPROTO_IP protocol value passed to socket(2) classify the socket as TCP or UDP. The patch also drops the check for SOCK_RAW and converts it into a default, so that socket types like SOCK_DCCP and SOCK_SEQPACKET are classified as SECCLASS_RAWIP_SOCKET (instead of generic sockets). Note that protocol-specific support for SCTP, DCCP etc. is not addressed here, we're just getting these protocols checked at the IP layer. This fixes a reported problem where SCTP sockets were being recognized as generic SELinux sockets yet still being passed in one case to an IP level check, which then fails for generic sockets. It will also fix bugs where any SOCK_STREAM socket is classified as TCP or any SOCK_DGRAM socket is classified as UDP. This patch also unifies the way IP sockets classes are determined in selinux_socket_bind(), so we use the already calculated value instead of trying to recalculate it. Signed-off-by: James Morris <jmorris@namei.org> Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-01 02:24:34 +08:00
static inline int default_protocol_stream(int protocol)
{
return (protocol == IPPROTO_IP || protocol == IPPROTO_TCP);
}
static inline int default_protocol_dgram(int protocol)
{
return (protocol == IPPROTO_IP || protocol == IPPROTO_UDP);
}
static inline u16 socket_type_to_security_class(int family, int type, int protocol)
{
switch (family) {
case PF_UNIX:
switch (type) {
case SOCK_STREAM:
case SOCK_SEQPACKET:
return SECCLASS_UNIX_STREAM_SOCKET;
case SOCK_DGRAM:
return SECCLASS_UNIX_DGRAM_SOCKET;
}
break;
case PF_INET:
case PF_INET6:
switch (type) {
case SOCK_STREAM:
[PATCH] SELinux - fix SCTP socket bug and general IP protocol handling The following patch updates the way SELinux classifies and handles IP based protocols. Currently, IP sockets are classified by SELinux as being either TCP, UDP or 'Raw', the latter being a default for IP socket that is not TCP or UDP. The classification code is out of date and uses only the socket type parameter to socket(2) to determine the class of IP socket. So, any socket created with SOCK_STREAM will be classified by SELinux as TCP, and SOCK_DGRAM as UDP. Also, other socket types such as SOCK_SEQPACKET and SOCK_DCCP are currently ignored by SELinux, which classifies them as generic sockets, which means they don't even get basic IP level checking. This patch changes the SELinux IP socket classification logic, so that only an IPPROTO_IP protocol value passed to socket(2) classify the socket as TCP or UDP. The patch also drops the check for SOCK_RAW and converts it into a default, so that socket types like SOCK_DCCP and SOCK_SEQPACKET are classified as SECCLASS_RAWIP_SOCKET (instead of generic sockets). Note that protocol-specific support for SCTP, DCCP etc. is not addressed here, we're just getting these protocols checked at the IP layer. This fixes a reported problem where SCTP sockets were being recognized as generic SELinux sockets yet still being passed in one case to an IP level check, which then fails for generic sockets. It will also fix bugs where any SOCK_STREAM socket is classified as TCP or any SOCK_DGRAM socket is classified as UDP. This patch also unifies the way IP sockets classes are determined in selinux_socket_bind(), so we use the already calculated value instead of trying to recalculate it. Signed-off-by: James Morris <jmorris@namei.org> Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-01 02:24:34 +08:00
if (default_protocol_stream(protocol))
return SECCLASS_TCP_SOCKET;
else
return SECCLASS_RAWIP_SOCKET;
case SOCK_DGRAM:
[PATCH] SELinux - fix SCTP socket bug and general IP protocol handling The following patch updates the way SELinux classifies and handles IP based protocols. Currently, IP sockets are classified by SELinux as being either TCP, UDP or 'Raw', the latter being a default for IP socket that is not TCP or UDP. The classification code is out of date and uses only the socket type parameter to socket(2) to determine the class of IP socket. So, any socket created with SOCK_STREAM will be classified by SELinux as TCP, and SOCK_DGRAM as UDP. Also, other socket types such as SOCK_SEQPACKET and SOCK_DCCP are currently ignored by SELinux, which classifies them as generic sockets, which means they don't even get basic IP level checking. This patch changes the SELinux IP socket classification logic, so that only an IPPROTO_IP protocol value passed to socket(2) classify the socket as TCP or UDP. The patch also drops the check for SOCK_RAW and converts it into a default, so that socket types like SOCK_DCCP and SOCK_SEQPACKET are classified as SECCLASS_RAWIP_SOCKET (instead of generic sockets). Note that protocol-specific support for SCTP, DCCP etc. is not addressed here, we're just getting these protocols checked at the IP layer. This fixes a reported problem where SCTP sockets were being recognized as generic SELinux sockets yet still being passed in one case to an IP level check, which then fails for generic sockets. It will also fix bugs where any SOCK_STREAM socket is classified as TCP or any SOCK_DGRAM socket is classified as UDP. This patch also unifies the way IP sockets classes are determined in selinux_socket_bind(), so we use the already calculated value instead of trying to recalculate it. Signed-off-by: James Morris <jmorris@namei.org> Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-01 02:24:34 +08:00
if (default_protocol_dgram(protocol))
return SECCLASS_UDP_SOCKET;
else
return SECCLASS_RAWIP_SOCKET;
case SOCK_DCCP:
return SECCLASS_DCCP_SOCKET;
[PATCH] SELinux - fix SCTP socket bug and general IP protocol handling The following patch updates the way SELinux classifies and handles IP based protocols. Currently, IP sockets are classified by SELinux as being either TCP, UDP or 'Raw', the latter being a default for IP socket that is not TCP or UDP. The classification code is out of date and uses only the socket type parameter to socket(2) to determine the class of IP socket. So, any socket created with SOCK_STREAM will be classified by SELinux as TCP, and SOCK_DGRAM as UDP. Also, other socket types such as SOCK_SEQPACKET and SOCK_DCCP are currently ignored by SELinux, which classifies them as generic sockets, which means they don't even get basic IP level checking. This patch changes the SELinux IP socket classification logic, so that only an IPPROTO_IP protocol value passed to socket(2) classify the socket as TCP or UDP. The patch also drops the check for SOCK_RAW and converts it into a default, so that socket types like SOCK_DCCP and SOCK_SEQPACKET are classified as SECCLASS_RAWIP_SOCKET (instead of generic sockets). Note that protocol-specific support for SCTP, DCCP etc. is not addressed here, we're just getting these protocols checked at the IP layer. This fixes a reported problem where SCTP sockets were being recognized as generic SELinux sockets yet still being passed in one case to an IP level check, which then fails for generic sockets. It will also fix bugs where any SOCK_STREAM socket is classified as TCP or any SOCK_DGRAM socket is classified as UDP. This patch also unifies the way IP sockets classes are determined in selinux_socket_bind(), so we use the already calculated value instead of trying to recalculate it. Signed-off-by: James Morris <jmorris@namei.org> Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-01 02:24:34 +08:00
default:
return SECCLASS_RAWIP_SOCKET;
}
break;
case PF_NETLINK:
switch (protocol) {
case NETLINK_ROUTE:
return SECCLASS_NETLINK_ROUTE_SOCKET;
case NETLINK_FIREWALL:
return SECCLASS_NETLINK_FIREWALL_SOCKET;
case NETLINK_INET_DIAG:
return SECCLASS_NETLINK_TCPDIAG_SOCKET;
case NETLINK_NFLOG:
return SECCLASS_NETLINK_NFLOG_SOCKET;
case NETLINK_XFRM:
return SECCLASS_NETLINK_XFRM_SOCKET;
case NETLINK_SELINUX:
return SECCLASS_NETLINK_SELINUX_SOCKET;
case NETLINK_AUDIT:
return SECCLASS_NETLINK_AUDIT_SOCKET;
case NETLINK_IP6_FW:
return SECCLASS_NETLINK_IP6FW_SOCKET;
case NETLINK_DNRTMSG:
return SECCLASS_NETLINK_DNRT_SOCKET;
case NETLINK_KOBJECT_UEVENT:
return SECCLASS_NETLINK_KOBJECT_UEVENT_SOCKET;
default:
return SECCLASS_NETLINK_SOCKET;
}
case PF_PACKET:
return SECCLASS_PACKET_SOCKET;
case PF_KEY:
return SECCLASS_KEY_SOCKET;
case PF_APPLETALK:
return SECCLASS_APPLETALK_SOCKET;
}
return SECCLASS_SOCKET;
}
#ifdef CONFIG_PROC_FS
static int selinux_proc_get_sid(struct proc_dir_entry *de,
u16 tclass,
u32 *sid)
{
int buflen, rc;
char *buffer, *path, *end;
buffer = (char *)__get_free_page(GFP_KERNEL);
if (!buffer)
return -ENOMEM;
buflen = PAGE_SIZE;
end = buffer+buflen;
*--end = '\0';
buflen--;
path = end-1;
*path = '/';
while (de && de != de->parent) {
buflen -= de->namelen + 1;
if (buflen < 0)
break;
end -= de->namelen;
memcpy(end, de->name, de->namelen);
*--end = '/';
path = end;
de = de->parent;
}
rc = security_genfs_sid("proc", path, tclass, sid);
free_page((unsigned long)buffer);
return rc;
}
#else
static int selinux_proc_get_sid(struct proc_dir_entry *de,
u16 tclass,
u32 *sid)
{
return -EINVAL;
}
#endif
/* The inode's security attributes must be initialized before first use. */
static int inode_doinit_with_dentry(struct inode *inode, struct dentry *opt_dentry)
{
struct superblock_security_struct *sbsec = NULL;
struct inode_security_struct *isec = inode->i_security;
u32 sid;
struct dentry *dentry;
#define INITCONTEXTLEN 255
char *context = NULL;
unsigned len = 0;
int rc = 0;
if (isec->initialized)
goto out;
mutex_lock(&isec->lock);
if (isec->initialized)
goto out_unlock;
sbsec = inode->i_sb->s_security;
if (!sbsec->initialized) {
/* Defer initialization until selinux_complete_init,
after the initial policy is loaded and the security
server is ready to handle calls. */
spin_lock(&sbsec->isec_lock);
if (list_empty(&isec->list))
list_add(&isec->list, &sbsec->isec_head);
spin_unlock(&sbsec->isec_lock);
goto out_unlock;
}
switch (sbsec->behavior) {
case SECURITY_FS_USE_XATTR:
if (!inode->i_op->getxattr) {
isec->sid = sbsec->def_sid;
break;
}
/* Need a dentry, since the xattr API requires one.
Life would be simpler if we could just pass the inode. */
if (opt_dentry) {
/* Called from d_instantiate or d_splice_alias. */
dentry = dget(opt_dentry);
} else {
/* Called from selinux_complete_init, try to find a dentry. */
dentry = d_find_alias(inode);
}
if (!dentry) {
printk(KERN_WARNING "SELinux: %s: no dentry for dev=%s "
"ino=%ld\n", __func__, inode->i_sb->s_id,
inode->i_ino);
goto out_unlock;
}
len = INITCONTEXTLEN;
context = kmalloc(len, GFP_NOFS);
if (!context) {
rc = -ENOMEM;
dput(dentry);
goto out_unlock;
}
rc = inode->i_op->getxattr(dentry, XATTR_NAME_SELINUX,
context, len);
if (rc == -ERANGE) {
/* Need a larger buffer. Query for the right size. */
rc = inode->i_op->getxattr(dentry, XATTR_NAME_SELINUX,
NULL, 0);
if (rc < 0) {
dput(dentry);
goto out_unlock;
}
kfree(context);
len = rc;
context = kmalloc(len, GFP_NOFS);
if (!context) {
rc = -ENOMEM;
dput(dentry);
goto out_unlock;
}
rc = inode->i_op->getxattr(dentry,
XATTR_NAME_SELINUX,
context, len);
}
dput(dentry);
if (rc < 0) {
if (rc != -ENODATA) {
printk(KERN_WARNING "SELinux: %s: getxattr returned "
"%d for dev=%s ino=%ld\n", __func__,
-rc, inode->i_sb->s_id, inode->i_ino);
kfree(context);
goto out_unlock;
}
/* Map ENODATA to the default file SID */
sid = sbsec->def_sid;
rc = 0;
} else {
rc = security_context_to_sid_default(context, rc, &sid,
sbsec->def_sid,
GFP_NOFS);
if (rc) {
printk(KERN_WARNING "SELinux: %s: context_to_sid(%s) "
"returned %d for dev=%s ino=%ld\n",
__func__, context, -rc,
inode->i_sb->s_id, inode->i_ino);
kfree(context);
/* Leave with the unlabeled SID */
rc = 0;
break;
}
}
kfree(context);
isec->sid = sid;
break;
case SECURITY_FS_USE_TASK:
isec->sid = isec->task_sid;
break;
case SECURITY_FS_USE_TRANS:
/* Default to the fs SID. */
isec->sid = sbsec->sid;
/* Try to obtain a transition SID. */
isec->sclass = inode_mode_to_security_class(inode->i_mode);
rc = security_transition_sid(isec->task_sid,
sbsec->sid,
isec->sclass,
&sid);
if (rc)
goto out_unlock;
isec->sid = sid;
break;
case SECURITY_FS_USE_MNTPOINT:
isec->sid = sbsec->mntpoint_sid;
break;
default:
/* Default to the fs superblock SID. */
isec->sid = sbsec->sid;
if (sbsec->proc && !S_ISLNK(inode->i_mode)) {
struct proc_inode *proci = PROC_I(inode);
if (proci->pde) {
isec->sclass = inode_mode_to_security_class(inode->i_mode);
rc = selinux_proc_get_sid(proci->pde,
isec->sclass,
&sid);
if (rc)
goto out_unlock;
isec->sid = sid;
}
}
break;
}
isec->initialized = 1;
out_unlock:
mutex_unlock(&isec->lock);
out:
if (isec->sclass == SECCLASS_FILE)
isec->sclass = inode_mode_to_security_class(inode->i_mode);
return rc;
}
/* Convert a Linux signal to an access vector. */
static inline u32 signal_to_av(int sig)
{
u32 perm = 0;
switch (sig) {
case SIGCHLD:
/* Commonly granted from child to parent. */
perm = PROCESS__SIGCHLD;
break;
case SIGKILL:
/* Cannot be caught or ignored */
perm = PROCESS__SIGKILL;
break;
case SIGSTOP:
/* Cannot be caught or ignored */
perm = PROCESS__SIGSTOP;
break;
default:
/* All other signals. */
perm = PROCESS__SIGNAL;
break;
}
return perm;
}
/* Check permission betweeen a pair of tasks, e.g. signal checks,
fork check, ptrace check, etc. */
static int task_has_perm(struct task_struct *tsk1,
struct task_struct *tsk2,
u32 perms)
{
struct task_security_struct *tsec1, *tsec2;
tsec1 = tsk1->security;
tsec2 = tsk2->security;
return avc_has_perm(tsec1->sid, tsec2->sid,
SECCLASS_PROCESS, perms, NULL);
}
#if CAP_LAST_CAP > 63
#error Fix SELinux to handle capabilities > 63.
#endif
/* Check whether a task is allowed to use a capability. */
static int task_has_capability(struct task_struct *tsk,
int cap)
{
struct task_security_struct *tsec;
struct avc_audit_data ad;
u16 sclass;
u32 av = CAP_TO_MASK(cap);
tsec = tsk->security;
AVC_AUDIT_DATA_INIT(&ad, CAP);
ad.tsk = tsk;
ad.u.cap = cap;
switch (CAP_TO_INDEX(cap)) {
case 0:
sclass = SECCLASS_CAPABILITY;
break;
case 1:
sclass = SECCLASS_CAPABILITY2;
break;
default:
printk(KERN_ERR
"SELinux: out of range capability %d\n", cap);
BUG();
}
return avc_has_perm(tsec->sid, tsec->sid, sclass, av, &ad);
}
/* Check whether a task is allowed to use a system operation. */
static int task_has_system(struct task_struct *tsk,
u32 perms)
{
struct task_security_struct *tsec;
tsec = tsk->security;
return avc_has_perm(tsec->sid, SECINITSID_KERNEL,
SECCLASS_SYSTEM, perms, NULL);
}
/* Check whether a task has a particular permission to an inode.
The 'adp' parameter is optional and allows other audit
data to be passed (e.g. the dentry). */
static int inode_has_perm(struct task_struct *tsk,
struct inode *inode,
u32 perms,
struct avc_audit_data *adp)
{
struct task_security_struct *tsec;
struct inode_security_struct *isec;
struct avc_audit_data ad;
if (unlikely(IS_PRIVATE(inode)))
return 0;
tsec = tsk->security;
isec = inode->i_security;
if (!adp) {
adp = &ad;
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.inode = inode;
}
return avc_has_perm(tsec->sid, isec->sid, isec->sclass, perms, adp);
}
/* Same as inode_has_perm, but pass explicit audit data containing
the dentry to help the auditing code to more easily generate the
pathname if needed. */
static inline int dentry_has_perm(struct task_struct *tsk,
struct vfsmount *mnt,
struct dentry *dentry,
u32 av)
{
struct inode *inode = dentry->d_inode;
struct avc_audit_data ad;
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.path.mnt = mnt;
ad.u.fs.path.dentry = dentry;
return inode_has_perm(tsk, inode, av, &ad);
}
/* Check whether a task can use an open file descriptor to
access an inode in a given way. Check access to the
descriptor itself, and then use dentry_has_perm to
check a particular permission to the file.
Access to the descriptor is implicitly granted if it
has the same SID as the process. If av is zero, then
access to the file is not checked, e.g. for cases
where only the descriptor is affected like seek. */
static int file_has_perm(struct task_struct *tsk,
struct file *file,
u32 av)
{
struct task_security_struct *tsec = tsk->security;
struct file_security_struct *fsec = file->f_security;
struct inode *inode = file->f_path.dentry->d_inode;
struct avc_audit_data ad;
int rc;
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.path = file->f_path;
if (tsec->sid != fsec->sid) {
rc = avc_has_perm(tsec->sid, fsec->sid,
SECCLASS_FD,
FD__USE,
&ad);
if (rc)
return rc;
}
/* av is zero if only checking access to the descriptor. */
if (av)
return inode_has_perm(tsk, inode, av, &ad);
return 0;
}
/* Check whether a task can create a file. */
static int may_create(struct inode *dir,
struct dentry *dentry,
u16 tclass)
{
struct task_security_struct *tsec;
struct inode_security_struct *dsec;
struct superblock_security_struct *sbsec;
u32 newsid;
struct avc_audit_data ad;
int rc;
tsec = current->security;
dsec = dir->i_security;
sbsec = dir->i_sb->s_security;
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.path.dentry = dentry;
rc = avc_has_perm(tsec->sid, dsec->sid, SECCLASS_DIR,
DIR__ADD_NAME | DIR__SEARCH,
&ad);
if (rc)
return rc;
if (tsec->create_sid && sbsec->behavior != SECURITY_FS_USE_MNTPOINT) {
newsid = tsec->create_sid;
} else {
rc = security_transition_sid(tsec->sid, dsec->sid, tclass,
&newsid);
if (rc)
return rc;
}
rc = avc_has_perm(tsec->sid, newsid, tclass, FILE__CREATE, &ad);
if (rc)
return rc;
return avc_has_perm(newsid, sbsec->sid,
SECCLASS_FILESYSTEM,
FILESYSTEM__ASSOCIATE, &ad);
}
/* Check whether a task can create a key. */
static int may_create_key(u32 ksid,
struct task_struct *ctx)
{
struct task_security_struct *tsec;
tsec = ctx->security;
return avc_has_perm(tsec->sid, ksid, SECCLASS_KEY, KEY__CREATE, NULL);
}
#define MAY_LINK 0
#define MAY_UNLINK 1
#define MAY_RMDIR 2
/* Check whether a task can link, unlink, or rmdir a file/directory. */
static int may_link(struct inode *dir,
struct dentry *dentry,
int kind)
{
struct task_security_struct *tsec;
struct inode_security_struct *dsec, *isec;
struct avc_audit_data ad;
u32 av;
int rc;
tsec = current->security;
dsec = dir->i_security;
isec = dentry->d_inode->i_security;
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.path.dentry = dentry;
av = DIR__SEARCH;
av |= (kind ? DIR__REMOVE_NAME : DIR__ADD_NAME);
rc = avc_has_perm(tsec->sid, dsec->sid, SECCLASS_DIR, av, &ad);
if (rc)
return rc;
switch (kind) {
case MAY_LINK:
av = FILE__LINK;
break;
case MAY_UNLINK:
av = FILE__UNLINK;
break;
case MAY_RMDIR:
av = DIR__RMDIR;
break;
default:
printk(KERN_WARNING "SELinux: %s: unrecognized kind %d\n",
__func__, kind);
return 0;
}
rc = avc_has_perm(tsec->sid, isec->sid, isec->sclass, av, &ad);
return rc;
}
static inline int may_rename(struct inode *old_dir,
struct dentry *old_dentry,
struct inode *new_dir,
struct dentry *new_dentry)
{
struct task_security_struct *tsec;
struct inode_security_struct *old_dsec, *new_dsec, *old_isec, *new_isec;
struct avc_audit_data ad;
u32 av;
int old_is_dir, new_is_dir;
int rc;
tsec = current->security;
old_dsec = old_dir->i_security;
old_isec = old_dentry->d_inode->i_security;
old_is_dir = S_ISDIR(old_dentry->d_inode->i_mode);
new_dsec = new_dir->i_security;
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.path.dentry = old_dentry;
rc = avc_has_perm(tsec->sid, old_dsec->sid, SECCLASS_DIR,
DIR__REMOVE_NAME | DIR__SEARCH, &ad);
if (rc)
return rc;
rc = avc_has_perm(tsec->sid, old_isec->sid,
old_isec->sclass, FILE__RENAME, &ad);
if (rc)
return rc;
if (old_is_dir && new_dir != old_dir) {
rc = avc_has_perm(tsec->sid, old_isec->sid,
old_isec->sclass, DIR__REPARENT, &ad);
if (rc)
return rc;
}
ad.u.fs.path.dentry = new_dentry;
av = DIR__ADD_NAME | DIR__SEARCH;
if (new_dentry->d_inode)
av |= DIR__REMOVE_NAME;
rc = avc_has_perm(tsec->sid, new_dsec->sid, SECCLASS_DIR, av, &ad);
if (rc)
return rc;
if (new_dentry->d_inode) {
new_isec = new_dentry->d_inode->i_security;
new_is_dir = S_ISDIR(new_dentry->d_inode->i_mode);
rc = avc_has_perm(tsec->sid, new_isec->sid,
new_isec->sclass,
(new_is_dir ? DIR__RMDIR : FILE__UNLINK), &ad);
if (rc)
return rc;
}
return 0;
}
/* Check whether a task can perform a filesystem operation. */
static int superblock_has_perm(struct task_struct *tsk,
struct super_block *sb,
u32 perms,
struct avc_audit_data *ad)
{
struct task_security_struct *tsec;
struct superblock_security_struct *sbsec;
tsec = tsk->security;
sbsec = sb->s_security;
return avc_has_perm(tsec->sid, sbsec->sid, SECCLASS_FILESYSTEM,
perms, ad);
}
/* Convert a Linux mode and permission mask to an access vector. */
static inline u32 file_mask_to_av(int mode, int mask)
{
u32 av = 0;
if ((mode & S_IFMT) != S_IFDIR) {
if (mask & MAY_EXEC)
av |= FILE__EXECUTE;
if (mask & MAY_READ)
av |= FILE__READ;
if (mask & MAY_APPEND)
av |= FILE__APPEND;
else if (mask & MAY_WRITE)
av |= FILE__WRITE;
} else {
if (mask & MAY_EXEC)
av |= DIR__SEARCH;
if (mask & MAY_WRITE)
av |= DIR__WRITE;
if (mask & MAY_READ)
av |= DIR__READ;
}
return av;
}
/*
* Convert a file mask to an access vector and include the correct open
* open permission.
*/
static inline u32 open_file_mask_to_av(int mode, int mask)
{
u32 av = file_mask_to_av(mode, mask);
if (selinux_policycap_openperm) {
/*
* lnk files and socks do not really have an 'open'
*/
if (S_ISREG(mode))
av |= FILE__OPEN;
else if (S_ISCHR(mode))
av |= CHR_FILE__OPEN;
else if (S_ISBLK(mode))
av |= BLK_FILE__OPEN;
else if (S_ISFIFO(mode))
av |= FIFO_FILE__OPEN;
else if (S_ISDIR(mode))
av |= DIR__OPEN;
else
printk(KERN_ERR "SELinux: WARNING: inside %s with "
"unknown mode:%x\n", __func__, mode);
}
return av;
}
/* Convert a Linux file to an access vector. */
static inline u32 file_to_av(struct file *file)
{
u32 av = 0;
if (file->f_mode & FMODE_READ)
av |= FILE__READ;
if (file->f_mode & FMODE_WRITE) {
if (file->f_flags & O_APPEND)
av |= FILE__APPEND;
else
av |= FILE__WRITE;
}
if (!av) {
/*
* Special file opened with flags 3 for ioctl-only use.
*/
av = FILE__IOCTL;
}
return av;
}
/* Hook functions begin here. */
security: Fix setting of PF_SUPERPRIV by __capable() Fix the setting of PF_SUPERPRIV by __capable() as it could corrupt the flags the target process if that is not the current process and it is trying to change its own flags in a different way at the same time. __capable() is using neither atomic ops nor locking to protect t->flags. This patch removes __capable() and introduces has_capability() that doesn't set PF_SUPERPRIV on the process being queried. This patch further splits security_ptrace() in two: (1) security_ptrace_may_access(). This passes judgement on whether one process may access another only (PTRACE_MODE_ATTACH for ptrace() and PTRACE_MODE_READ for /proc), and takes a pointer to the child process. current is the parent. (2) security_ptrace_traceme(). This passes judgement on PTRACE_TRACEME only, and takes only a pointer to the parent process. current is the child. In Smack and commoncap, this uses has_capability() to determine whether the parent will be permitted to use PTRACE_ATTACH if normal checks fail. This does not set PF_SUPERPRIV. Two of the instances of __capable() actually only act on current, and so have been changed to calls to capable(). Of the places that were using __capable(): (1) The OOM killer calls __capable() thrice when weighing the killability of a process. All of these now use has_capability(). (2) cap_ptrace() and smack_ptrace() were using __capable() to check to see whether the parent was allowed to trace any process. As mentioned above, these have been split. For PTRACE_ATTACH and /proc, capable() is now used, and for PTRACE_TRACEME, has_capability() is used. (3) cap_safe_nice() only ever saw current, so now uses capable(). (4) smack_setprocattr() rejected accesses to tasks other than current just after calling __capable(), so the order of these two tests have been switched and capable() is used instead. (5) In smack_file_send_sigiotask(), we need to allow privileged processes to receive SIGIO on files they're manipulating. (6) In smack_task_wait(), we let a process wait for a privileged process, whether or not the process doing the waiting is privileged. I've tested this with the LTP SELinux and syscalls testscripts. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Serge Hallyn <serue@us.ibm.com> Acked-by: Casey Schaufler <casey@schaufler-ca.com> Acked-by: Andrew G. Morgan <morgan@kernel.org> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: James Morris <jmorris@namei.org>
2008-08-14 18:37:28 +08:00
static int selinux_ptrace_may_access(struct task_struct *child,
unsigned int mode)
{
int rc;
security: Fix setting of PF_SUPERPRIV by __capable() Fix the setting of PF_SUPERPRIV by __capable() as it could corrupt the flags the target process if that is not the current process and it is trying to change its own flags in a different way at the same time. __capable() is using neither atomic ops nor locking to protect t->flags. This patch removes __capable() and introduces has_capability() that doesn't set PF_SUPERPRIV on the process being queried. This patch further splits security_ptrace() in two: (1) security_ptrace_may_access(). This passes judgement on whether one process may access another only (PTRACE_MODE_ATTACH for ptrace() and PTRACE_MODE_READ for /proc), and takes a pointer to the child process. current is the parent. (2) security_ptrace_traceme(). This passes judgement on PTRACE_TRACEME only, and takes only a pointer to the parent process. current is the child. In Smack and commoncap, this uses has_capability() to determine whether the parent will be permitted to use PTRACE_ATTACH if normal checks fail. This does not set PF_SUPERPRIV. Two of the instances of __capable() actually only act on current, and so have been changed to calls to capable(). Of the places that were using __capable(): (1) The OOM killer calls __capable() thrice when weighing the killability of a process. All of these now use has_capability(). (2) cap_ptrace() and smack_ptrace() were using __capable() to check to see whether the parent was allowed to trace any process. As mentioned above, these have been split. For PTRACE_ATTACH and /proc, capable() is now used, and for PTRACE_TRACEME, has_capability() is used. (3) cap_safe_nice() only ever saw current, so now uses capable(). (4) smack_setprocattr() rejected accesses to tasks other than current just after calling __capable(), so the order of these two tests have been switched and capable() is used instead. (5) In smack_file_send_sigiotask(), we need to allow privileged processes to receive SIGIO on files they're manipulating. (6) In smack_task_wait(), we let a process wait for a privileged process, whether or not the process doing the waiting is privileged. I've tested this with the LTP SELinux and syscalls testscripts. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Serge Hallyn <serue@us.ibm.com> Acked-by: Casey Schaufler <casey@schaufler-ca.com> Acked-by: Andrew G. Morgan <morgan@kernel.org> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: James Morris <jmorris@namei.org>
2008-08-14 18:37:28 +08:00
rc = secondary_ops->ptrace_may_access(child, mode);
if (rc)
return rc;
Security: split proc ptrace checking into read vs. attach Enable security modules to distinguish reading of process state via proc from full ptrace access by renaming ptrace_may_attach to ptrace_may_access and adding a mode argument indicating whether only read access or full attach access is requested. This allows security modules to permit access to reading process state without granting full ptrace access. The base DAC/capability checking remains unchanged. Read access to /proc/pid/mem continues to apply a full ptrace attach check since check_mem_permission() already requires the current task to already be ptracing the target. The other ptrace checks within proc for elements like environ, maps, and fds are changed to pass the read mode instead of attach. In the SELinux case, we model such reading of process state as a reading of a proc file labeled with the target process' label. This enables SELinux policy to permit such reading of process state without permitting control or manipulation of the target process, as there are a number of cases where programs probe for such information via proc but do not need to be able to control the target (e.g. procps, lsof, PolicyKit, ConsoleKit). At present we have to choose between allowing full ptrace in policy (more permissive than required/desired) or breaking functionality (or in some cases just silencing the denials via dontaudit rules but this can hide genuine attacks). This version of the patch incorporates comments from Casey Schaufler (change/replace existing ptrace_may_attach interface, pass access mode), and Chris Wright (provide greater consistency in the checking). Note that like their predecessors __ptrace_may_attach and ptrace_may_attach, the __ptrace_may_access and ptrace_may_access interfaces use different return value conventions from each other (0 or -errno vs. 1 or 0). I retained this difference to avoid any changes to the caller logic but made the difference clearer by changing the latter interface to return a bool rather than an int and by adding a comment about it to ptrace.h for any future callers. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Acked-by: Chris Wright <chrisw@sous-sol.org> Signed-off-by: James Morris <jmorris@namei.org>
2008-05-19 20:32:49 +08:00
if (mode == PTRACE_MODE_READ) {
security: Fix setting of PF_SUPERPRIV by __capable() Fix the setting of PF_SUPERPRIV by __capable() as it could corrupt the flags the target process if that is not the current process and it is trying to change its own flags in a different way at the same time. __capable() is using neither atomic ops nor locking to protect t->flags. This patch removes __capable() and introduces has_capability() that doesn't set PF_SUPERPRIV on the process being queried. This patch further splits security_ptrace() in two: (1) security_ptrace_may_access(). This passes judgement on whether one process may access another only (PTRACE_MODE_ATTACH for ptrace() and PTRACE_MODE_READ for /proc), and takes a pointer to the child process. current is the parent. (2) security_ptrace_traceme(). This passes judgement on PTRACE_TRACEME only, and takes only a pointer to the parent process. current is the child. In Smack and commoncap, this uses has_capability() to determine whether the parent will be permitted to use PTRACE_ATTACH if normal checks fail. This does not set PF_SUPERPRIV. Two of the instances of __capable() actually only act on current, and so have been changed to calls to capable(). Of the places that were using __capable(): (1) The OOM killer calls __capable() thrice when weighing the killability of a process. All of these now use has_capability(). (2) cap_ptrace() and smack_ptrace() were using __capable() to check to see whether the parent was allowed to trace any process. As mentioned above, these have been split. For PTRACE_ATTACH and /proc, capable() is now used, and for PTRACE_TRACEME, has_capability() is used. (3) cap_safe_nice() only ever saw current, so now uses capable(). (4) smack_setprocattr() rejected accesses to tasks other than current just after calling __capable(), so the order of these two tests have been switched and capable() is used instead. (5) In smack_file_send_sigiotask(), we need to allow privileged processes to receive SIGIO on files they're manipulating. (6) In smack_task_wait(), we let a process wait for a privileged process, whether or not the process doing the waiting is privileged. I've tested this with the LTP SELinux and syscalls testscripts. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Serge Hallyn <serue@us.ibm.com> Acked-by: Casey Schaufler <casey@schaufler-ca.com> Acked-by: Andrew G. Morgan <morgan@kernel.org> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: James Morris <jmorris@namei.org>
2008-08-14 18:37:28 +08:00
struct task_security_struct *tsec = current->security;
Security: split proc ptrace checking into read vs. attach Enable security modules to distinguish reading of process state via proc from full ptrace access by renaming ptrace_may_attach to ptrace_may_access and adding a mode argument indicating whether only read access or full attach access is requested. This allows security modules to permit access to reading process state without granting full ptrace access. The base DAC/capability checking remains unchanged. Read access to /proc/pid/mem continues to apply a full ptrace attach check since check_mem_permission() already requires the current task to already be ptracing the target. The other ptrace checks within proc for elements like environ, maps, and fds are changed to pass the read mode instead of attach. In the SELinux case, we model such reading of process state as a reading of a proc file labeled with the target process' label. This enables SELinux policy to permit such reading of process state without permitting control or manipulation of the target process, as there are a number of cases where programs probe for such information via proc but do not need to be able to control the target (e.g. procps, lsof, PolicyKit, ConsoleKit). At present we have to choose between allowing full ptrace in policy (more permissive than required/desired) or breaking functionality (or in some cases just silencing the denials via dontaudit rules but this can hide genuine attacks). This version of the patch incorporates comments from Casey Schaufler (change/replace existing ptrace_may_attach interface, pass access mode), and Chris Wright (provide greater consistency in the checking). Note that like their predecessors __ptrace_may_attach and ptrace_may_attach, the __ptrace_may_access and ptrace_may_access interfaces use different return value conventions from each other (0 or -errno vs. 1 or 0). I retained this difference to avoid any changes to the caller logic but made the difference clearer by changing the latter interface to return a bool rather than an int and by adding a comment about it to ptrace.h for any future callers. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Acked-by: Chris Wright <chrisw@sous-sol.org> Signed-off-by: James Morris <jmorris@namei.org>
2008-05-19 20:32:49 +08:00
struct task_security_struct *csec = child->security;
return avc_has_perm(tsec->sid, csec->sid,
SECCLASS_FILE, FILE__READ, NULL);
}
security: Fix setting of PF_SUPERPRIV by __capable() Fix the setting of PF_SUPERPRIV by __capable() as it could corrupt the flags the target process if that is not the current process and it is trying to change its own flags in a different way at the same time. __capable() is using neither atomic ops nor locking to protect t->flags. This patch removes __capable() and introduces has_capability() that doesn't set PF_SUPERPRIV on the process being queried. This patch further splits security_ptrace() in two: (1) security_ptrace_may_access(). This passes judgement on whether one process may access another only (PTRACE_MODE_ATTACH for ptrace() and PTRACE_MODE_READ for /proc), and takes a pointer to the child process. current is the parent. (2) security_ptrace_traceme(). This passes judgement on PTRACE_TRACEME only, and takes only a pointer to the parent process. current is the child. In Smack and commoncap, this uses has_capability() to determine whether the parent will be permitted to use PTRACE_ATTACH if normal checks fail. This does not set PF_SUPERPRIV. Two of the instances of __capable() actually only act on current, and so have been changed to calls to capable(). Of the places that were using __capable(): (1) The OOM killer calls __capable() thrice when weighing the killability of a process. All of these now use has_capability(). (2) cap_ptrace() and smack_ptrace() were using __capable() to check to see whether the parent was allowed to trace any process. As mentioned above, these have been split. For PTRACE_ATTACH and /proc, capable() is now used, and for PTRACE_TRACEME, has_capability() is used. (3) cap_safe_nice() only ever saw current, so now uses capable(). (4) smack_setprocattr() rejected accesses to tasks other than current just after calling __capable(), so the order of these two tests have been switched and capable() is used instead. (5) In smack_file_send_sigiotask(), we need to allow privileged processes to receive SIGIO on files they're manipulating. (6) In smack_task_wait(), we let a process wait for a privileged process, whether or not the process doing the waiting is privileged. I've tested this with the LTP SELinux and syscalls testscripts. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Serge Hallyn <serue@us.ibm.com> Acked-by: Casey Schaufler <casey@schaufler-ca.com> Acked-by: Andrew G. Morgan <morgan@kernel.org> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: James Morris <jmorris@namei.org>
2008-08-14 18:37:28 +08:00
return task_has_perm(current, child, PROCESS__PTRACE);
}
static int selinux_ptrace_traceme(struct task_struct *parent)
{
int rc;
rc = secondary_ops->ptrace_traceme(parent);
if (rc)
return rc;
return task_has_perm(parent, current, PROCESS__PTRACE);
}
static int selinux_capget(struct task_struct *target, kernel_cap_t *effective,
kernel_cap_t *inheritable, kernel_cap_t *permitted)
{
int error;
error = task_has_perm(current, target, PROCESS__GETCAP);
if (error)
return error;
return secondary_ops->capget(target, effective, inheritable, permitted);
}
static int selinux_capset_check(struct task_struct *target, kernel_cap_t *effective,
kernel_cap_t *inheritable, kernel_cap_t *permitted)
{
int error;
error = secondary_ops->capset_check(target, effective, inheritable, permitted);
if (error)
return error;
return task_has_perm(current, target, PROCESS__SETCAP);
}
static void selinux_capset_set(struct task_struct *target, kernel_cap_t *effective,
kernel_cap_t *inheritable, kernel_cap_t *permitted)
{
secondary_ops->capset_set(target, effective, inheritable, permitted);
}
static int selinux_capable(struct task_struct *tsk, int cap)
{
int rc;
rc = secondary_ops->capable(tsk, cap);
if (rc)
return rc;
return task_has_capability(tsk, cap);
}
static int selinux_sysctl_get_sid(ctl_table *table, u16 tclass, u32 *sid)
{
int buflen, rc;
char *buffer, *path, *end;
rc = -ENOMEM;
buffer = (char *)__get_free_page(GFP_KERNEL);
if (!buffer)
goto out;
buflen = PAGE_SIZE;
end = buffer+buflen;
*--end = '\0';
buflen--;
path = end-1;
*path = '/';
while (table) {
const char *name = table->procname;
size_t namelen = strlen(name);
buflen -= namelen + 1;
if (buflen < 0)
goto out_free;
end -= namelen;
memcpy(end, name, namelen);
*--end = '/';
path = end;
table = table->parent;
}
buflen -= 4;
if (buflen < 0)
goto out_free;
end -= 4;
memcpy(end, "/sys", 4);
path = end;
rc = security_genfs_sid("proc", path, tclass, sid);
out_free:
free_page((unsigned long)buffer);
out:
return rc;
}
static int selinux_sysctl(ctl_table *table, int op)
{
int error = 0;
u32 av;
struct task_security_struct *tsec;
u32 tsid;
int rc;
rc = secondary_ops->sysctl(table, op);
if (rc)
return rc;
tsec = current->security;
rc = selinux_sysctl_get_sid(table, (op == 0001) ?
SECCLASS_DIR : SECCLASS_FILE, &tsid);
if (rc) {
/* Default to the well-defined sysctl SID. */
tsid = SECINITSID_SYSCTL;
}
/* The op values are "defined" in sysctl.c, thereby creating
* a bad coupling between this module and sysctl.c */
if (op == 001) {
error = avc_has_perm(tsec->sid, tsid,
SECCLASS_DIR, DIR__SEARCH, NULL);
} else {
av = 0;
if (op & 004)
av |= FILE__READ;
if (op & 002)
av |= FILE__WRITE;
if (av)
error = avc_has_perm(tsec->sid, tsid,
SECCLASS_FILE, av, NULL);
}
return error;
}
static int selinux_quotactl(int cmds, int type, int id, struct super_block *sb)
{
int rc = 0;
if (!sb)
return 0;
switch (cmds) {
case Q_SYNC:
case Q_QUOTAON:
case Q_QUOTAOFF:
case Q_SETINFO:
case Q_SETQUOTA:
rc = superblock_has_perm(current, sb, FILESYSTEM__QUOTAMOD,
NULL);
break;
case Q_GETFMT:
case Q_GETINFO:
case Q_GETQUOTA:
rc = superblock_has_perm(current, sb, FILESYSTEM__QUOTAGET,
NULL);
break;
default:
rc = 0; /* let the kernel handle invalid cmds */
break;
}
return rc;
}
static int selinux_quota_on(struct dentry *dentry)
{
return dentry_has_perm(current, NULL, dentry, FILE__QUOTAON);
}
static int selinux_syslog(int type)
{
int rc;
rc = secondary_ops->syslog(type);
if (rc)
return rc;
switch (type) {
case 3: /* Read last kernel messages */
case 10: /* Return size of the log buffer */
rc = task_has_system(current, SYSTEM__SYSLOG_READ);
break;
case 6: /* Disable logging to console */
case 7: /* Enable logging to console */
case 8: /* Set level of messages printed to console */
rc = task_has_system(current, SYSTEM__SYSLOG_CONSOLE);
break;
case 0: /* Close log */
case 1: /* Open log */
case 2: /* Read from log */
case 4: /* Read/clear last kernel messages */
case 5: /* Clear ring buffer */
default:
rc = task_has_system(current, SYSTEM__SYSLOG_MOD);
break;
}
return rc;
}
/*
* Check that a process has enough memory to allocate a new virtual
* mapping. 0 means there is enough memory for the allocation to
* succeed and -ENOMEM implies there is not.
*
* Note that secondary_ops->capable and task_has_perm_noaudit return 0
* if the capability is granted, but __vm_enough_memory requires 1 if
* the capability is granted.
*
* Do not audit the selinux permission check, as this is applied to all
* processes that allocate mappings.
*/
static int selinux_vm_enough_memory(struct mm_struct *mm, long pages)
{
int rc, cap_sys_admin = 0;
struct task_security_struct *tsec = current->security;
rc = secondary_ops->capable(current, CAP_SYS_ADMIN);
if (rc == 0)
rc = avc_has_perm_noaudit(tsec->sid, tsec->sid,
SECCLASS_CAPABILITY,
CAP_TO_MASK(CAP_SYS_ADMIN),
0,
NULL);
if (rc == 0)
cap_sys_admin = 1;
return __vm_enough_memory(mm, pages, cap_sys_admin);
}
/* binprm security operations */
static int selinux_bprm_alloc_security(struct linux_binprm *bprm)
{
struct bprm_security_struct *bsec;
bsec = kzalloc(sizeof(struct bprm_security_struct), GFP_KERNEL);
if (!bsec)
return -ENOMEM;
bsec->sid = SECINITSID_UNLABELED;
bsec->set = 0;
bprm->security = bsec;
return 0;
}
static int selinux_bprm_set_security(struct linux_binprm *bprm)
{
struct task_security_struct *tsec;
struct inode *inode = bprm->file->f_path.dentry->d_inode;
struct inode_security_struct *isec;
struct bprm_security_struct *bsec;
u32 newsid;
struct avc_audit_data ad;
int rc;
rc = secondary_ops->bprm_set_security(bprm);
if (rc)
return rc;
bsec = bprm->security;
if (bsec->set)
return 0;
tsec = current->security;
isec = inode->i_security;
/* Default to the current task SID. */
bsec->sid = tsec->sid;
/* Reset fs, key, and sock SIDs on execve. */
tsec->create_sid = 0;
tsec->keycreate_sid = 0;
tsec->sockcreate_sid = 0;
if (tsec->exec_sid) {
newsid = tsec->exec_sid;
/* Reset exec SID on execve. */
tsec->exec_sid = 0;
} else {
/* Check for a default transition on this program. */
rc = security_transition_sid(tsec->sid, isec->sid,
SECCLASS_PROCESS, &newsid);
if (rc)
return rc;
}
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.path = bprm->file->f_path;
if (bprm->file->f_path.mnt->mnt_flags & MNT_NOSUID)
newsid = tsec->sid;
if (tsec->sid == newsid) {
rc = avc_has_perm(tsec->sid, isec->sid,
SECCLASS_FILE, FILE__EXECUTE_NO_TRANS, &ad);
if (rc)
return rc;
} else {
/* Check permissions for the transition. */
rc = avc_has_perm(tsec->sid, newsid,
SECCLASS_PROCESS, PROCESS__TRANSITION, &ad);
if (rc)
return rc;
rc = avc_has_perm(newsid, isec->sid,
SECCLASS_FILE, FILE__ENTRYPOINT, &ad);
if (rc)
return rc;
/* Clear any possibly unsafe personality bits on exec: */
current->personality &= ~PER_CLEAR_ON_SETID;
/* Set the security field to the new SID. */
bsec->sid = newsid;
}
bsec->set = 1;
return 0;
}
static int selinux_bprm_check_security(struct linux_binprm *bprm)
{
return secondary_ops->bprm_check_security(bprm);
}
static int selinux_bprm_secureexec(struct linux_binprm *bprm)
{
struct task_security_struct *tsec = current->security;
int atsecure = 0;
if (tsec->osid != tsec->sid) {
/* Enable secure mode for SIDs transitions unless
the noatsecure permission is granted between
the two SIDs, i.e. ahp returns 0. */
atsecure = avc_has_perm(tsec->osid, tsec->sid,
SECCLASS_PROCESS,
PROCESS__NOATSECURE, NULL);
}
return (atsecure || secondary_ops->bprm_secureexec(bprm));
}
static void selinux_bprm_free_security(struct linux_binprm *bprm)
{
kfree(bprm->security);
bprm->security = NULL;
}
extern struct vfsmount *selinuxfs_mount;
extern struct dentry *selinux_null;
/* Derived from fs/exec.c:flush_old_files. */
static inline void flush_unauthorized_files(struct files_struct *files)
{
struct avc_audit_data ad;
struct file *file, *devnull = NULL;
struct tty_struct *tty;
struct fdtable *fdt;
long j = -1;
[PATCH] tty: ->signal->tty locking Fix the locking of signal->tty. Use ->sighand->siglock to protect ->signal->tty; this lock is already used by most other members of ->signal/->sighand. And unless we are 'current' or the tasklist_lock is held we need ->siglock to access ->signal anyway. (NOTE: sys_unshare() is broken wrt ->sighand locking rules) Note that tty_mutex is held over tty destruction, so while holding tty_mutex any tty pointer remains valid. Otherwise the lifetime of ttys are governed by their open file handles. This leaves some holes for tty access from signal->tty (or any other non file related tty access). It solves the tty SLAB scribbles we were seeing. (NOTE: the change from group_send_sig_info to __group_send_sig_info needs to be examined by someone familiar with the security framework, I think it is safe given the SEND_SIG_PRIV from other __group_send_sig_info invocations) [schwidefsky@de.ibm.com: 3270 fix] [akpm@osdl.org: various post-viro fixes] Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Alan Cox <alan@redhat.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Cc: Prarit Bhargava <prarit@redhat.com> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Roland McGrath <roland@redhat.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Jan Kara <jack@ucw.cz> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-08 18:36:04 +08:00
int drop_tty = 0;
[PATCH] tty: ->signal->tty locking Fix the locking of signal->tty. Use ->sighand->siglock to protect ->signal->tty; this lock is already used by most other members of ->signal/->sighand. And unless we are 'current' or the tasklist_lock is held we need ->siglock to access ->signal anyway. (NOTE: sys_unshare() is broken wrt ->sighand locking rules) Note that tty_mutex is held over tty destruction, so while holding tty_mutex any tty pointer remains valid. Otherwise the lifetime of ttys are governed by their open file handles. This leaves some holes for tty access from signal->tty (or any other non file related tty access). It solves the tty SLAB scribbles we were seeing. (NOTE: the change from group_send_sig_info to __group_send_sig_info needs to be examined by someone familiar with the security framework, I think it is safe given the SEND_SIG_PRIV from other __group_send_sig_info invocations) [schwidefsky@de.ibm.com: 3270 fix] [akpm@osdl.org: various post-viro fixes] Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Alan Cox <alan@redhat.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Cc: Prarit Bhargava <prarit@redhat.com> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Roland McGrath <roland@redhat.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Jan Kara <jack@ucw.cz> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-08 18:36:04 +08:00
tty = get_current_tty();
if (tty) {
file_list_lock();
file = list_entry(tty->tty_files.next, typeof(*file), f_u.fu_list);
if (file) {
/* Revalidate access to controlling tty.
Use inode_has_perm on the tty inode directly rather
than using file_has_perm, as this particular open
file may belong to another process and we are only
interested in the inode-based check here. */
struct inode *inode = file->f_path.dentry->d_inode;
if (inode_has_perm(current, inode,
FILE__READ | FILE__WRITE, NULL)) {
[PATCH] tty: ->signal->tty locking Fix the locking of signal->tty. Use ->sighand->siglock to protect ->signal->tty; this lock is already used by most other members of ->signal/->sighand. And unless we are 'current' or the tasklist_lock is held we need ->siglock to access ->signal anyway. (NOTE: sys_unshare() is broken wrt ->sighand locking rules) Note that tty_mutex is held over tty destruction, so while holding tty_mutex any tty pointer remains valid. Otherwise the lifetime of ttys are governed by their open file handles. This leaves some holes for tty access from signal->tty (or any other non file related tty access). It solves the tty SLAB scribbles we were seeing. (NOTE: the change from group_send_sig_info to __group_send_sig_info needs to be examined by someone familiar with the security framework, I think it is safe given the SEND_SIG_PRIV from other __group_send_sig_info invocations) [schwidefsky@de.ibm.com: 3270 fix] [akpm@osdl.org: various post-viro fixes] Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl> Acked-by: Alan Cox <alan@redhat.com> Cc: Oleg Nesterov <oleg@tv-sign.ru> Cc: Prarit Bhargava <prarit@redhat.com> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Roland McGrath <roland@redhat.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Jan Kara <jack@ucw.cz> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-12-08 18:36:04 +08:00
drop_tty = 1;
}
}
file_list_unlock();
tty_kref_put(tty);
}
/* Reset controlling tty. */
if (drop_tty)
no_tty();
/* Revalidate access to inherited open files. */
AVC_AUDIT_DATA_INIT(&ad, FS);
spin_lock(&files->file_lock);
for (;;) {
unsigned long set, i;
int fd;
j++;
i = j * __NFDBITS;
fdt = files_fdtable(files);
if (i >= fdt->max_fds)
break;
set = fdt->open_fds->fds_bits[j];
if (!set)
continue;
spin_unlock(&files->file_lock);
for ( ; set ; i++, set >>= 1) {
if (set & 1) {
file = fget(i);
if (!file)
continue;
if (file_has_perm(current,
file,
file_to_av(file))) {
sys_close(i);
fd = get_unused_fd();
if (fd != i) {
if (fd >= 0)
put_unused_fd(fd);
fput(file);
continue;
}
if (devnull) {
get_file(devnull);
} else {
devnull = dentry_open(dget(selinux_null), mntget(selinuxfs_mount), O_RDWR);
if (IS_ERR(devnull)) {
devnull = NULL;
put_unused_fd(fd);
fput(file);
continue;
}
}
fd_install(fd, devnull);
}
fput(file);
}
}
spin_lock(&files->file_lock);
}
spin_unlock(&files->file_lock);
}
static void selinux_bprm_apply_creds(struct linux_binprm *bprm, int unsafe)
{
struct task_security_struct *tsec;
struct bprm_security_struct *bsec;
u32 sid;
int rc;
secondary_ops->bprm_apply_creds(bprm, unsafe);
tsec = current->security;
bsec = bprm->security;
sid = bsec->sid;
tsec->osid = tsec->sid;
bsec->unsafe = 0;
if (tsec->sid != sid) {
/* Check for shared state. If not ok, leave SID
unchanged and kill. */
if (unsafe & LSM_UNSAFE_SHARE) {
rc = avc_has_perm(tsec->sid, sid, SECCLASS_PROCESS,
PROCESS__SHARE, NULL);
if (rc) {
bsec->unsafe = 1;
return;
}
}
/* Check for ptracing, and update the task SID if ok.
Otherwise, leave SID unchanged and kill. */
if (unsafe & (LSM_UNSAFE_PTRACE | LSM_UNSAFE_PTRACE_CAP)) {
struct task_struct *tracer;
struct task_security_struct *sec;
u32 ptsid = 0;
rcu_read_lock();
tracer = tracehook_tracer_task(current);
if (likely(tracer != NULL)) {
sec = tracer->security;
ptsid = sec->sid;
}
rcu_read_unlock();
if (ptsid != 0) {
rc = avc_has_perm(ptsid, sid, SECCLASS_PROCESS,
PROCESS__PTRACE, NULL);
if (rc) {
bsec->unsafe = 1;
return;
}
}
}
tsec->sid = sid;
}
}
/*
* called after apply_creds without the task lock held
*/
static void selinux_bprm_post_apply_creds(struct linux_binprm *bprm)
{
struct task_security_struct *tsec;
struct rlimit *rlim, *initrlim;
struct itimerval itimer;
struct bprm_security_struct *bsec;
int rc, i;
tsec = current->security;
bsec = bprm->security;
if (bsec->unsafe) {
force_sig_specific(SIGKILL, current);
return;
}
if (tsec->osid == tsec->sid)
return;
/* Close files for which the new task SID is not authorized. */
flush_unauthorized_files(current->files);
/* Check whether the new SID can inherit signal state
from the old SID. If not, clear itimers to avoid
subsequent signal generation and flush and unblock
signals. This must occur _after_ the task SID has
been updated so that any kill done after the flush
will be checked against the new SID. */
rc = avc_has_perm(tsec->osid, tsec->sid, SECCLASS_PROCESS,
PROCESS__SIGINH, NULL);
if (rc) {
memset(&itimer, 0, sizeof itimer);
for (i = 0; i < 3; i++)
do_setitimer(i, &itimer, NULL);
flush_signals(current);
spin_lock_irq(&current->sighand->siglock);
flush_signal_handlers(current, 1);
sigemptyset(&current->blocked);
recalc_sigpending();
spin_unlock_irq(&current->sighand->siglock);
}
/* Always clear parent death signal on SID transitions. */
current->pdeath_signal = 0;
/* Check whether the new SID can inherit resource limits
from the old SID. If not, reset all soft limits to
the lower of the current task's hard limit and the init
task's soft limit. Note that the setting of hard limits
(even to lower them) can be controlled by the setrlimit
check. The inclusion of the init task's soft limit into
the computation is to avoid resetting soft limits higher
than the default soft limit for cases where the default
is lower than the hard limit, e.g. RLIMIT_CORE or
RLIMIT_STACK.*/
rc = avc_has_perm(tsec->osid, tsec->sid, SECCLASS_PROCESS,
PROCESS__RLIMITINH, NULL);
if (rc) {
for (i = 0; i < RLIM_NLIMITS; i++) {
rlim = current->signal->rlim + i;
initrlim = init_task.signal->rlim+i;
rlim->rlim_cur = min(rlim->rlim_max, initrlim->rlim_cur);
}
timers: fix itimer/many thread hang Overview This patch reworks the handling of POSIX CPU timers, including the ITIMER_PROF, ITIMER_VIRT timers and rlimit handling. It was put together with the help of Roland McGrath, the owner and original writer of this code. The problem we ran into, and the reason for this rework, has to do with using a profiling timer in a process with a large number of threads. It appears that the performance of the old implementation of run_posix_cpu_timers() was at least O(n*3) (where "n" is the number of threads in a process) or worse. Everything is fine with an increasing number of threads until the time taken for that routine to run becomes the same as or greater than the tick time, at which point things degrade rather quickly. This patch fixes bug 9906, "Weird hang with NPTL and SIGPROF." Code Changes This rework corrects the implementation of run_posix_cpu_timers() to make it run in constant time for a particular machine. (Performance may vary between one machine and another depending upon whether the kernel is built as single- or multiprocessor and, in the latter case, depending upon the number of running processors.) To do this, at each tick we now update fields in signal_struct as well as task_struct. The run_posix_cpu_timers() function uses those fields to make its decisions. We define a new structure, "task_cputime," to contain user, system and scheduler times and use these in appropriate places: struct task_cputime { cputime_t utime; cputime_t stime; unsigned long long sum_exec_runtime; }; This is included in the structure "thread_group_cputime," which is a new substructure of signal_struct and which varies for uniprocessor versus multiprocessor kernels. For uniprocessor kernels, it uses "task_cputime" as a simple substructure, while for multiprocessor kernels it is a pointer: struct thread_group_cputime { struct task_cputime totals; }; struct thread_group_cputime { struct task_cputime *totals; }; We also add a new task_cputime substructure directly to signal_struct, to cache the earliest expiration of process-wide timers, and task_cputime also replaces the it_*_expires fields of task_struct (used for earliest expiration of thread timers). The "thread_group_cputime" structure contains process-wide timers that are updated via account_user_time() and friends. In the non-SMP case the structure is a simple aggregator; unfortunately in the SMP case that simplicity was not achievable due to cache-line contention between CPUs (in one measured case performance was actually _worse_ on a 16-cpu system than the same test on a 4-cpu system, due to this contention). For SMP, the thread_group_cputime counters are maintained as a per-cpu structure allocated using alloc_percpu(). The timer functions update only the timer field in the structure corresponding to the running CPU, obtained using per_cpu_ptr(). We define a set of inline functions in sched.h that we use to maintain the thread_group_cputime structure and hide the differences between UP and SMP implementations from the rest of the kernel. The thread_group_cputime_init() function initializes the thread_group_cputime structure for the given task. The thread_group_cputime_alloc() is a no-op for UP; for SMP it calls the out-of-line function thread_group_cputime_alloc_smp() to allocate and fill in the per-cpu structures and fields. The thread_group_cputime_free() function, also a no-op for UP, in SMP frees the per-cpu structures. The thread_group_cputime_clone_thread() function (also a UP no-op) for SMP calls thread_group_cputime_alloc() if the per-cpu structures haven't yet been allocated. The thread_group_cputime() function fills the task_cputime structure it is passed with the contents of the thread_group_cputime fields; in UP it's that simple but in SMP it must also safely check that tsk->signal is non-NULL (if it is it just uses the appropriate fields of task_struct) and, if so, sums the per-cpu values for each online CPU. Finally, the three functions account_group_user_time(), account_group_system_time() and account_group_exec_runtime() are used by timer functions to update the respective fields of the thread_group_cputime structure. Non-SMP operation is trivial and will not be mentioned further. The per-cpu structure is always allocated when a task creates its first new thread, via a call to thread_group_cputime_clone_thread() from copy_signal(). It is freed at process exit via a call to thread_group_cputime_free() from cleanup_signal(). All functions that formerly summed utime/stime/sum_sched_runtime values from from all threads in the thread group now use thread_group_cputime() to snapshot the values in the thread_group_cputime structure or the values in the task structure itself if the per-cpu structure hasn't been allocated. Finally, the code in kernel/posix-cpu-timers.c has changed quite a bit. The run_posix_cpu_timers() function has been split into a fast path and a slow path; the former safely checks whether there are any expired thread timers and, if not, just returns, while the slow path does the heavy lifting. With the dedicated thread group fields, timers are no longer "rebalanced" and the process_timer_rebalance() function and related code has gone away. All summing loops are gone and all code that used them now uses the thread_group_cputime() inline. When process-wide timers are set, the new task_cputime structure in signal_struct is used to cache the earliest expiration; this is checked in the fast path. Performance The fix appears not to add significant overhead to existing operations. It generally performs the same as the current code except in two cases, one in which it performs slightly worse (Case 5 below) and one in which it performs very significantly better (Case 2 below). Overall it's a wash except in those two cases. I've since done somewhat more involved testing on a dual-core Opteron system. Case 1: With no itimer running, for a test with 100,000 threads, the fixed kernel took 1428.5 seconds, 513 seconds more than the unfixed system, all of which was spent in the system. There were twice as many voluntary context switches with the fix as without it. Case 2: With an itimer running at .01 second ticks and 4000 threads (the most an unmodified kernel can handle), the fixed kernel ran the test in eight percent of the time (5.8 seconds as opposed to 70 seconds) and had better tick accuracy (.012 seconds per tick as opposed to .023 seconds per tick). Case 3: A 4000-thread test with an initial timer tick of .01 second and an interval of 10,000 seconds (i.e. a timer that ticks only once) had very nearly the same performance in both cases: 6.3 seconds elapsed for the fixed kernel versus 5.5 seconds for the unfixed kernel. With fewer threads (eight in these tests), the Case 1 test ran in essentially the same time on both the modified and unmodified kernels (5.2 seconds versus 5.8 seconds). The Case 2 test ran in about the same time as well, 5.9 seconds versus 5.4 seconds but again with much better tick accuracy, .013 seconds per tick versus .025 seconds per tick for the unmodified kernel. Since the fix affected the rlimit code, I also tested soft and hard CPU limits. Case 4: With a hard CPU limit of 20 seconds and eight threads (and an itimer running), the modified kernel was very slightly favored in that while it killed the process in 19.997 seconds of CPU time (5.002 seconds of wall time), only .003 seconds of that was system time, the rest was user time. The unmodified kernel killed the process in 20.001 seconds of CPU (5.014 seconds of wall time) of which .016 seconds was system time. Really, though, the results were too close to call. The results were essentially the same with no itimer running. Case 5: With a soft limit of 20 seconds and a hard limit of 2000 seconds (where the hard limit would never be reached) and an itimer running, the modified kernel exhibited worse tick accuracy than the unmodified kernel: .050 seconds/tick versus .028 seconds/tick. Otherwise, performance was almost indistinguishable. With no itimer running this test exhibited virtually identical behavior and times in both cases. In times past I did some limited performance testing. those results are below. On a four-cpu Opteron system without this fix, a sixteen-thread test executed in 3569.991 seconds, of which user was 3568.435s and system was 1.556s. On the same system with the fix, user and elapsed time were about the same, but system time dropped to 0.007 seconds. Performance with eight, four and one thread were comparable. Interestingly, the timer ticks with the fix seemed more accurate: The sixteen-thread test with the fix received 149543 ticks for 0.024 seconds per tick, while the same test without the fix received 58720 for 0.061 seconds per tick. Both cases were configured for an interval of 0.01 seconds. Again, the other tests were comparable. Each thread in this test computed the primes up to 25,000,000. I also did a test with a large number of threads, 100,000 threads, which is impossible without the fix. In this case each thread computed the primes only up to 10,000 (to make the runtime manageable). System time dominated, at 1546.968 seconds out of a total 2176.906 seconds (giving a user time of 629.938s). It received 147651 ticks for 0.015 seconds per tick, still quite accurate. There is obviously no comparable test without the fix. Signed-off-by: Frank Mayhar <fmayhar@google.com> Cc: Roland McGrath <roland@redhat.com> Cc: Alexey Dobriyan <adobriyan@gmail.com> Cc: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-09-13 00:54:39 +08:00
update_rlimit_cpu(rlim->rlim_cur);
}
/* Wake up the parent if it is waiting so that it can
recheck wait permission to the new task SID. */
wake_up_interruptible(&current->parent->signal->wait_chldexit);
}
/* superblock security operations */
static int selinux_sb_alloc_security(struct super_block *sb)
{
return superblock_alloc_security(sb);
}
static void selinux_sb_free_security(struct super_block *sb)
{
superblock_free_security(sb);
}
static inline int match_prefix(char *prefix, int plen, char *option, int olen)
{
if (plen > olen)
return 0;
return !memcmp(prefix, option, plen);
}
static inline int selinux_option(char *option, int len)
{
return (match_prefix(CONTEXT_STR, sizeof(CONTEXT_STR)-1, option, len) ||
match_prefix(FSCONTEXT_STR, sizeof(FSCONTEXT_STR)-1, option, len) ||
match_prefix(DEFCONTEXT_STR, sizeof(DEFCONTEXT_STR)-1, option, len) ||
match_prefix(ROOTCONTEXT_STR, sizeof(ROOTCONTEXT_STR)-1, option, len));
}
static inline void take_option(char **to, char *from, int *first, int len)
{
if (!*first) {
**to = ',';
*to += 1;
} else
*first = 0;
memcpy(*to, from, len);
*to += len;
}
static inline void take_selinux_option(char **to, char *from, int *first,
int len)
{
int current_size = 0;
if (!*first) {
**to = '|';
*to += 1;
} else
*first = 0;
while (current_size < len) {
if (*from != '"') {
**to = *from;
*to += 1;
}
from += 1;
current_size += 1;
}
}
static int selinux_sb_copy_data(char *orig, char *copy)
{
int fnosec, fsec, rc = 0;
char *in_save, *in_curr, *in_end;
char *sec_curr, *nosec_save, *nosec;
int open_quote = 0;
in_curr = orig;
sec_curr = copy;
nosec = (char *)get_zeroed_page(GFP_KERNEL);
if (!nosec) {
rc = -ENOMEM;
goto out;
}
nosec_save = nosec;
fnosec = fsec = 1;
in_save = in_end = orig;
do {
if (*in_end == '"')
open_quote = !open_quote;
if ((*in_end == ',' && open_quote == 0) ||
*in_end == '\0') {
int len = in_end - in_curr;
if (selinux_option(in_curr, len))
take_selinux_option(&sec_curr, in_curr, &fsec, len);
else
take_option(&nosec, in_curr, &fnosec, len);
in_curr = in_end + 1;
}
} while (*in_end++);
strcpy(in_save, nosec_save);
free_page((unsigned long)nosec_save);
out:
return rc;
}
static int selinux_sb_kern_mount(struct super_block *sb, void *data)
{
struct avc_audit_data ad;
int rc;
rc = superblock_doinit(sb, data);
if (rc)
return rc;
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.path.dentry = sb->s_root;
return superblock_has_perm(current, sb, FILESYSTEM__MOUNT, &ad);
}
static int selinux_sb_statfs(struct dentry *dentry)
{
struct avc_audit_data ad;
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.path.dentry = dentry->d_sb->s_root;
return superblock_has_perm(current, dentry->d_sb, FILESYSTEM__GETATTR, &ad);
}
static int selinux_mount(char *dev_name,
struct path *path,
char *type,
unsigned long flags,
void *data)
{
int rc;
rc = secondary_ops->sb_mount(dev_name, path, type, flags, data);
if (rc)
return rc;
if (flags & MS_REMOUNT)
return superblock_has_perm(current, path->mnt->mnt_sb,
FILESYSTEM__REMOUNT, NULL);
else
return dentry_has_perm(current, path->mnt, path->dentry,
FILE__MOUNTON);
}
static int selinux_umount(struct vfsmount *mnt, int flags)
{
int rc;
rc = secondary_ops->sb_umount(mnt, flags);
if (rc)
return rc;
return superblock_has_perm(current, mnt->mnt_sb,
FILESYSTEM__UNMOUNT, NULL);
}
/* inode security operations */
static int selinux_inode_alloc_security(struct inode *inode)
{
return inode_alloc_security(inode);
}
static void selinux_inode_free_security(struct inode *inode)
{
inode_free_security(inode);
}
[PATCH] security: enable atomic inode security labeling The following patch set enables atomic security labeling of newly created inodes by altering the fs code to invoke a new LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state during the inode creation transaction. This parallels the existing processing for setting ACLs on newly created inodes. Otherwise, it is possible for new inodes to be accessed by another thread via the dcache prior to complete security setup (presently handled by the post_create/mkdir/... LSM hooks in the VFS) and a newly created inode may be left unlabeled on the disk in the event of a crash. SELinux presently works around the issue by ensuring that the incore inode security label is initialized to a special SID that is inaccessible to unprivileged processes (in accordance with policy), thereby preventing inappropriate access but potentially causing false denials on legitimate accesses. A simple test program demonstrates such false denials on SELinux, and the patch solves the problem. Similar such false denials have been encountered in real applications. This patch defines a new inode_init_security LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state for it, and adds a corresponding hook function implementation to SELinux. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-09-10 04:01:35 +08:00
static int selinux_inode_init_security(struct inode *inode, struct inode *dir,
char **name, void **value,
size_t *len)
{
struct task_security_struct *tsec;
struct inode_security_struct *dsec;
struct superblock_security_struct *sbsec;
u32 newsid, clen;
[PATCH] security: enable atomic inode security labeling The following patch set enables atomic security labeling of newly created inodes by altering the fs code to invoke a new LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state during the inode creation transaction. This parallels the existing processing for setting ACLs on newly created inodes. Otherwise, it is possible for new inodes to be accessed by another thread via the dcache prior to complete security setup (presently handled by the post_create/mkdir/... LSM hooks in the VFS) and a newly created inode may be left unlabeled on the disk in the event of a crash. SELinux presently works around the issue by ensuring that the incore inode security label is initialized to a special SID that is inaccessible to unprivileged processes (in accordance with policy), thereby preventing inappropriate access but potentially causing false denials on legitimate accesses. A simple test program demonstrates such false denials on SELinux, and the patch solves the problem. Similar such false denials have been encountered in real applications. This patch defines a new inode_init_security LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state for it, and adds a corresponding hook function implementation to SELinux. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-09-10 04:01:35 +08:00
int rc;
char *namep = NULL, *context;
[PATCH] security: enable atomic inode security labeling The following patch set enables atomic security labeling of newly created inodes by altering the fs code to invoke a new LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state during the inode creation transaction. This parallels the existing processing for setting ACLs on newly created inodes. Otherwise, it is possible for new inodes to be accessed by another thread via the dcache prior to complete security setup (presently handled by the post_create/mkdir/... LSM hooks in the VFS) and a newly created inode may be left unlabeled on the disk in the event of a crash. SELinux presently works around the issue by ensuring that the incore inode security label is initialized to a special SID that is inaccessible to unprivileged processes (in accordance with policy), thereby preventing inappropriate access but potentially causing false denials on legitimate accesses. A simple test program demonstrates such false denials on SELinux, and the patch solves the problem. Similar such false denials have been encountered in real applications. This patch defines a new inode_init_security LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state for it, and adds a corresponding hook function implementation to SELinux. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-09-10 04:01:35 +08:00
tsec = current->security;
dsec = dir->i_security;
sbsec = dir->i_sb->s_security;
if (tsec->create_sid && sbsec->behavior != SECURITY_FS_USE_MNTPOINT) {
newsid = tsec->create_sid;
} else {
rc = security_transition_sid(tsec->sid, dsec->sid,
inode_mode_to_security_class(inode->i_mode),
&newsid);
if (rc) {
printk(KERN_WARNING "%s: "
"security_transition_sid failed, rc=%d (dev=%s "
"ino=%ld)\n",
__func__,
[PATCH] security: enable atomic inode security labeling The following patch set enables atomic security labeling of newly created inodes by altering the fs code to invoke a new LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state during the inode creation transaction. This parallels the existing processing for setting ACLs on newly created inodes. Otherwise, it is possible for new inodes to be accessed by another thread via the dcache prior to complete security setup (presently handled by the post_create/mkdir/... LSM hooks in the VFS) and a newly created inode may be left unlabeled on the disk in the event of a crash. SELinux presently works around the issue by ensuring that the incore inode security label is initialized to a special SID that is inaccessible to unprivileged processes (in accordance with policy), thereby preventing inappropriate access but potentially causing false denials on legitimate accesses. A simple test program demonstrates such false denials on SELinux, and the patch solves the problem. Similar such false denials have been encountered in real applications. This patch defines a new inode_init_security LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state for it, and adds a corresponding hook function implementation to SELinux. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-09-10 04:01:35 +08:00
-rc, inode->i_sb->s_id, inode->i_ino);
return rc;
}
}
/* Possibly defer initialization to selinux_complete_init. */
if (sbsec->initialized) {
struct inode_security_struct *isec = inode->i_security;
isec->sclass = inode_mode_to_security_class(inode->i_mode);
isec->sid = newsid;
isec->initialized = 1;
}
[PATCH] security: enable atomic inode security labeling The following patch set enables atomic security labeling of newly created inodes by altering the fs code to invoke a new LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state during the inode creation transaction. This parallels the existing processing for setting ACLs on newly created inodes. Otherwise, it is possible for new inodes to be accessed by another thread via the dcache prior to complete security setup (presently handled by the post_create/mkdir/... LSM hooks in the VFS) and a newly created inode may be left unlabeled on the disk in the event of a crash. SELinux presently works around the issue by ensuring that the incore inode security label is initialized to a special SID that is inaccessible to unprivileged processes (in accordance with policy), thereby preventing inappropriate access but potentially causing false denials on legitimate accesses. A simple test program demonstrates such false denials on SELinux, and the patch solves the problem. Similar such false denials have been encountered in real applications. This patch defines a new inode_init_security LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state for it, and adds a corresponding hook function implementation to SELinux. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-09-10 04:01:35 +08:00
if (!ss_initialized || sbsec->behavior == SECURITY_FS_USE_MNTPOINT)
return -EOPNOTSUPP;
if (name) {
namep = kstrdup(XATTR_SELINUX_SUFFIX, GFP_NOFS);
if (!namep)
return -ENOMEM;
*name = namep;
}
[PATCH] security: enable atomic inode security labeling The following patch set enables atomic security labeling of newly created inodes by altering the fs code to invoke a new LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state during the inode creation transaction. This parallels the existing processing for setting ACLs on newly created inodes. Otherwise, it is possible for new inodes to be accessed by another thread via the dcache prior to complete security setup (presently handled by the post_create/mkdir/... LSM hooks in the VFS) and a newly created inode may be left unlabeled on the disk in the event of a crash. SELinux presently works around the issue by ensuring that the incore inode security label is initialized to a special SID that is inaccessible to unprivileged processes (in accordance with policy), thereby preventing inappropriate access but potentially causing false denials on legitimate accesses. A simple test program demonstrates such false denials on SELinux, and the patch solves the problem. Similar such false denials have been encountered in real applications. This patch defines a new inode_init_security LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state for it, and adds a corresponding hook function implementation to SELinux. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-09-10 04:01:35 +08:00
if (value && len) {
selinux: support deferred mapping of contexts Introduce SELinux support for deferred mapping of security contexts in the SID table upon policy reload, and use this support for inode security contexts when the context is not yet valid under the current policy. Only processes with CAP_MAC_ADMIN + mac_admin permission in policy can set undefined security contexts on inodes. Inodes with such undefined contexts are treated as having the unlabeled context until the context becomes valid upon a policy reload that defines the context. Context invalidation upon policy reload also uses this support to save the context information in the SID table and later recover it upon a subsequent policy reload that defines the context again. This support is to enable package managers and similar programs to set down file contexts unknown to the system policy at the time the file is created in order to better support placing loadable policy modules in packages and to support build systems that need to create images of different distro releases with different policies w/o requiring all of the contexts to be defined or legal in the build host policy. With this patch applied, the following sequence is possible, although in practice it is recommended that this permission only be allowed to specific program domains such as the package manager. # rmdir baz # rm bar # touch bar # chcon -t foo_exec_t bar # foo_exec_t is not yet defined chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument # cat setundefined.te policy_module(setundefined, 1.0) require { type unconfined_t; type unlabeled_t; } files_type(unlabeled_t) allow unconfined_t self:capability2 mac_admin; # make -f /usr/share/selinux/devel/Makefile setundefined.pp # semodule -i setundefined.pp # chcon -t foo_exec_t bar # foo_exec_t is not yet defined # mkdir -Z system_u:object_r:foo_exec_t baz # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # cat foo.te policy_module(foo, 1.0) type foo_exec_t; files_type(foo_exec_t) # make -f /usr/share/selinux/devel/Makefile foo.pp # semodule -i foo.pp # defines foo_exec_t # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r foo # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # semodule -i foo.pp # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r setundefined foo # chcon -t foo_exec_t bar # no longer defined and not allowed chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # rmdir baz # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2008-05-08 01:03:20 +08:00
rc = security_sid_to_context_force(newsid, &context, &clen);
if (rc) {
kfree(namep);
return rc;
}
*value = context;
*len = clen;
[PATCH] security: enable atomic inode security labeling The following patch set enables atomic security labeling of newly created inodes by altering the fs code to invoke a new LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state during the inode creation transaction. This parallels the existing processing for setting ACLs on newly created inodes. Otherwise, it is possible for new inodes to be accessed by another thread via the dcache prior to complete security setup (presently handled by the post_create/mkdir/... LSM hooks in the VFS) and a newly created inode may be left unlabeled on the disk in the event of a crash. SELinux presently works around the issue by ensuring that the incore inode security label is initialized to a special SID that is inaccessible to unprivileged processes (in accordance with policy), thereby preventing inappropriate access but potentially causing false denials on legitimate accesses. A simple test program demonstrates such false denials on SELinux, and the patch solves the problem. Similar such false denials have been encountered in real applications. This patch defines a new inode_init_security LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state for it, and adds a corresponding hook function implementation to SELinux. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-09-10 04:01:35 +08:00
}
return 0;
}
static int selinux_inode_create(struct inode *dir, struct dentry *dentry, int mask)
{
return may_create(dir, dentry, SECCLASS_FILE);
}
static int selinux_inode_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_dentry)
{
int rc;
rc = secondary_ops->inode_link(old_dentry, dir, new_dentry);
if (rc)
return rc;
return may_link(dir, old_dentry, MAY_LINK);
}
static int selinux_inode_unlink(struct inode *dir, struct dentry *dentry)
{
int rc;
rc = secondary_ops->inode_unlink(dir, dentry);
if (rc)
return rc;
return may_link(dir, dentry, MAY_UNLINK);
}
static int selinux_inode_symlink(struct inode *dir, struct dentry *dentry, const char *name)
{
return may_create(dir, dentry, SECCLASS_LNK_FILE);
}
static int selinux_inode_mkdir(struct inode *dir, struct dentry *dentry, int mask)
{
return may_create(dir, dentry, SECCLASS_DIR);
}
static int selinux_inode_rmdir(struct inode *dir, struct dentry *dentry)
{
return may_link(dir, dentry, MAY_RMDIR);
}
static int selinux_inode_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t dev)
{
int rc;
rc = secondary_ops->inode_mknod(dir, dentry, mode, dev);
if (rc)
return rc;
return may_create(dir, dentry, inode_mode_to_security_class(mode));
}
static int selinux_inode_rename(struct inode *old_inode, struct dentry *old_dentry,
struct inode *new_inode, struct dentry *new_dentry)
{
return may_rename(old_inode, old_dentry, new_inode, new_dentry);
}
static int selinux_inode_readlink(struct dentry *dentry)
{
return dentry_has_perm(current, NULL, dentry, FILE__READ);
}
static int selinux_inode_follow_link(struct dentry *dentry, struct nameidata *nameidata)
{
int rc;
rc = secondary_ops->inode_follow_link(dentry, nameidata);
if (rc)
return rc;
return dentry_has_perm(current, NULL, dentry, FILE__READ);
}
static int selinux_inode_permission(struct inode *inode, int mask)
{
int rc;
rc = secondary_ops->inode_permission(inode, mask);
if (rc)
return rc;
if (!mask) {
/* No permission to check. Existence test. */
return 0;
}
return inode_has_perm(current, inode,
open_file_mask_to_av(inode->i_mode, mask), NULL);
}
static int selinux_inode_setattr(struct dentry *dentry, struct iattr *iattr)
{
int rc;
rc = secondary_ops->inode_setattr(dentry, iattr);
if (rc)
return rc;
if (iattr->ia_valid & ATTR_FORCE)
return 0;
if (iattr->ia_valid & (ATTR_MODE | ATTR_UID | ATTR_GID |
ATTR_ATIME_SET | ATTR_MTIME_SET))
return dentry_has_perm(current, NULL, dentry, FILE__SETATTR);
return dentry_has_perm(current, NULL, dentry, FILE__WRITE);
}
static int selinux_inode_getattr(struct vfsmount *mnt, struct dentry *dentry)
{
return dentry_has_perm(current, mnt, dentry, FILE__GETATTR);
}
static int selinux_inode_setotherxattr(struct dentry *dentry, const char *name)
Implement file posix capabilities Implement file posix capabilities. This allows programs to be given a subset of root's powers regardless of who runs them, without having to use setuid and giving the binary all of root's powers. This version works with Kaigai Kohei's userspace tools, found at http://www.kaigai.gr.jp/index.php. For more information on how to use this patch, Chris Friedhoff has posted a nice page at http://www.friedhoff.org/fscaps.html. Changelog: Nov 27: Incorporate fixes from Andrew Morton (security-introduce-file-caps-tweaks and security-introduce-file-caps-warning-fix) Fix Kconfig dependency. Fix change signaling behavior when file caps are not compiled in. Nov 13: Integrate comments from Alexey: Remove CONFIG_ ifdef from capability.h, and use %zd for printing a size_t. Nov 13: Fix endianness warnings by sparse as suggested by Alexey Dobriyan. Nov 09: Address warnings of unused variables at cap_bprm_set_security when file capabilities are disabled, and simultaneously clean up the code a little, by pulling the new code into a helper function. Nov 08: For pointers to required userspace tools and how to use them, see http://www.friedhoff.org/fscaps.html. Nov 07: Fix the calculation of the highest bit checked in check_cap_sanity(). Nov 07: Allow file caps to be enabled without CONFIG_SECURITY, since capabilities are the default. Hook cap_task_setscheduler when !CONFIG_SECURITY. Move capable(TASK_KILL) to end of cap_task_kill to reduce audit messages. Nov 05: Add secondary calls in selinux/hooks.c to task_setioprio and task_setscheduler so that selinux and capabilities with file cap support can be stacked. Sep 05: As Seth Arnold points out, uid checks are out of place for capability code. Sep 01: Define task_setscheduler, task_setioprio, cap_task_kill, and task_setnice to make sure a user cannot affect a process in which they called a program with some fscaps. One remaining question is the note under task_setscheduler: are we ok with CAP_SYS_NICE being sufficient to confine a process to a cpuset? It is a semantic change, as without fsccaps, attach_task doesn't allow CAP_SYS_NICE to override the uid equivalence check. But since it uses security_task_setscheduler, which elsewhere is used where CAP_SYS_NICE can be used to override the uid equivalence check, fixing it might be tough. task_setscheduler note: this also controls cpuset:attach_task. Are we ok with CAP_SYS_NICE being used to confine to a cpuset? task_setioprio task_setnice sys_setpriority uses this (through set_one_prio) for another process. Need same checks as setrlimit Aug 21: Updated secureexec implementation to reflect the fact that euid and uid might be the same and nonzero, but the process might still have elevated caps. Aug 15: Handle endianness of xattrs. Enforce capability version match between kernel and disk. Enforce that no bits beyond the known max capability are set, else return -EPERM. With this extra processing, it may be worth reconsidering doing all the work at bprm_set_security rather than d_instantiate. Aug 10: Always call getxattr at bprm_set_security, rather than caching it at d_instantiate. [morgan@kernel.org: file-caps clean up for linux/capability.h] [bunk@kernel.org: unexport cap_inode_killpriv] Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Andrew Morgan <morgan@kernel.org> Signed-off-by: Andrew Morgan <morgan@kernel.org> Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 14:31:36 +08:00
{
if (!strncmp(name, XATTR_SECURITY_PREFIX,
sizeof XATTR_SECURITY_PREFIX - 1)) {
if (!strcmp(name, XATTR_NAME_CAPS)) {
if (!capable(CAP_SETFCAP))
return -EPERM;
} else if (!capable(CAP_SYS_ADMIN)) {
/* A different attribute in the security namespace.
Restrict to administrator. */
return -EPERM;
}
}
/* Not an attribute we recognize, so just check the
ordinary setattr permission. */
return dentry_has_perm(current, NULL, dentry, FILE__SETATTR);
}
static int selinux_inode_setxattr(struct dentry *dentry, const char *name,
const void *value, size_t size, int flags)
{
struct task_security_struct *tsec = current->security;
struct inode *inode = dentry->d_inode;
struct inode_security_struct *isec = inode->i_security;
struct superblock_security_struct *sbsec;
struct avc_audit_data ad;
u32 newsid;
int rc = 0;
Implement file posix capabilities Implement file posix capabilities. This allows programs to be given a subset of root's powers regardless of who runs them, without having to use setuid and giving the binary all of root's powers. This version works with Kaigai Kohei's userspace tools, found at http://www.kaigai.gr.jp/index.php. For more information on how to use this patch, Chris Friedhoff has posted a nice page at http://www.friedhoff.org/fscaps.html. Changelog: Nov 27: Incorporate fixes from Andrew Morton (security-introduce-file-caps-tweaks and security-introduce-file-caps-warning-fix) Fix Kconfig dependency. Fix change signaling behavior when file caps are not compiled in. Nov 13: Integrate comments from Alexey: Remove CONFIG_ ifdef from capability.h, and use %zd for printing a size_t. Nov 13: Fix endianness warnings by sparse as suggested by Alexey Dobriyan. Nov 09: Address warnings of unused variables at cap_bprm_set_security when file capabilities are disabled, and simultaneously clean up the code a little, by pulling the new code into a helper function. Nov 08: For pointers to required userspace tools and how to use them, see http://www.friedhoff.org/fscaps.html. Nov 07: Fix the calculation of the highest bit checked in check_cap_sanity(). Nov 07: Allow file caps to be enabled without CONFIG_SECURITY, since capabilities are the default. Hook cap_task_setscheduler when !CONFIG_SECURITY. Move capable(TASK_KILL) to end of cap_task_kill to reduce audit messages. Nov 05: Add secondary calls in selinux/hooks.c to task_setioprio and task_setscheduler so that selinux and capabilities with file cap support can be stacked. Sep 05: As Seth Arnold points out, uid checks are out of place for capability code. Sep 01: Define task_setscheduler, task_setioprio, cap_task_kill, and task_setnice to make sure a user cannot affect a process in which they called a program with some fscaps. One remaining question is the note under task_setscheduler: are we ok with CAP_SYS_NICE being sufficient to confine a process to a cpuset? It is a semantic change, as without fsccaps, attach_task doesn't allow CAP_SYS_NICE to override the uid equivalence check. But since it uses security_task_setscheduler, which elsewhere is used where CAP_SYS_NICE can be used to override the uid equivalence check, fixing it might be tough. task_setscheduler note: this also controls cpuset:attach_task. Are we ok with CAP_SYS_NICE being used to confine to a cpuset? task_setioprio task_setnice sys_setpriority uses this (through set_one_prio) for another process. Need same checks as setrlimit Aug 21: Updated secureexec implementation to reflect the fact that euid and uid might be the same and nonzero, but the process might still have elevated caps. Aug 15: Handle endianness of xattrs. Enforce capability version match between kernel and disk. Enforce that no bits beyond the known max capability are set, else return -EPERM. With this extra processing, it may be worth reconsidering doing all the work at bprm_set_security rather than d_instantiate. Aug 10: Always call getxattr at bprm_set_security, rather than caching it at d_instantiate. [morgan@kernel.org: file-caps clean up for linux/capability.h] [bunk@kernel.org: unexport cap_inode_killpriv] Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Andrew Morgan <morgan@kernel.org> Signed-off-by: Andrew Morgan <morgan@kernel.org> Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 14:31:36 +08:00
if (strcmp(name, XATTR_NAME_SELINUX))
return selinux_inode_setotherxattr(dentry, name);
sbsec = inode->i_sb->s_security;
if (sbsec->behavior == SECURITY_FS_USE_MNTPOINT)
return -EOPNOTSUPP;
if (!is_owner_or_cap(inode))
return -EPERM;
AVC_AUDIT_DATA_INIT(&ad, FS);
ad.u.fs.path.dentry = dentry;
rc = avc_has_perm(tsec->sid, isec->sid, isec->sclass,
FILE__RELABELFROM, &ad);
if (rc)
return rc;
rc = security_context_to_sid(value, size, &newsid);
selinux: support deferred mapping of contexts Introduce SELinux support for deferred mapping of security contexts in the SID table upon policy reload, and use this support for inode security contexts when the context is not yet valid under the current policy. Only processes with CAP_MAC_ADMIN + mac_admin permission in policy can set undefined security contexts on inodes. Inodes with such undefined contexts are treated as having the unlabeled context until the context becomes valid upon a policy reload that defines the context. Context invalidation upon policy reload also uses this support to save the context information in the SID table and later recover it upon a subsequent policy reload that defines the context again. This support is to enable package managers and similar programs to set down file contexts unknown to the system policy at the time the file is created in order to better support placing loadable policy modules in packages and to support build systems that need to create images of different distro releases with different policies w/o requiring all of the contexts to be defined or legal in the build host policy. With this patch applied, the following sequence is possible, although in practice it is recommended that this permission only be allowed to specific program domains such as the package manager. # rmdir baz # rm bar # touch bar # chcon -t foo_exec_t bar # foo_exec_t is not yet defined chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument # cat setundefined.te policy_module(setundefined, 1.0) require { type unconfined_t; type unlabeled_t; } files_type(unlabeled_t) allow unconfined_t self:capability2 mac_admin; # make -f /usr/share/selinux/devel/Makefile setundefined.pp # semodule -i setundefined.pp # chcon -t foo_exec_t bar # foo_exec_t is not yet defined # mkdir -Z system_u:object_r:foo_exec_t baz # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # cat foo.te policy_module(foo, 1.0) type foo_exec_t; files_type(foo_exec_t) # make -f /usr/share/selinux/devel/Makefile foo.pp # semodule -i foo.pp # defines foo_exec_t # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r foo # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # semodule -i foo.pp # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r setundefined foo # chcon -t foo_exec_t bar # no longer defined and not allowed chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # rmdir baz # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2008-05-08 01:03:20 +08:00
if (rc == -EINVAL) {
if (!capable(CAP_MAC_ADMIN))
return rc;
rc = security_context_to_sid_force(value, size, &newsid);
}
if (rc)
return rc;
rc = avc_has_perm(tsec->sid, newsid, isec->sclass,
FILE__RELABELTO, &ad);
if (rc)
return rc;
rc = security_validate_transition(isec->sid, newsid, tsec->sid,
isec->sclass);
if (rc)
return rc;
return avc_has_perm(newsid,
sbsec->sid,
SECCLASS_FILESYSTEM,
FILESYSTEM__ASSOCIATE,
&ad);
}
static void selinux_inode_post_setxattr(struct dentry *dentry, const char *name,
const void *value, size_t size,
int flags)
{
struct inode *inode = dentry->d_inode;
struct inode_security_struct *isec = inode->i_security;
u32 newsid;
int rc;
if (strcmp(name, XATTR_NAME_SELINUX)) {
/* Not an attribute we recognize, so nothing to do. */
return;
}
selinux: support deferred mapping of contexts Introduce SELinux support for deferred mapping of security contexts in the SID table upon policy reload, and use this support for inode security contexts when the context is not yet valid under the current policy. Only processes with CAP_MAC_ADMIN + mac_admin permission in policy can set undefined security contexts on inodes. Inodes with such undefined contexts are treated as having the unlabeled context until the context becomes valid upon a policy reload that defines the context. Context invalidation upon policy reload also uses this support to save the context information in the SID table and later recover it upon a subsequent policy reload that defines the context again. This support is to enable package managers and similar programs to set down file contexts unknown to the system policy at the time the file is created in order to better support placing loadable policy modules in packages and to support build systems that need to create images of different distro releases with different policies w/o requiring all of the contexts to be defined or legal in the build host policy. With this patch applied, the following sequence is possible, although in practice it is recommended that this permission only be allowed to specific program domains such as the package manager. # rmdir baz # rm bar # touch bar # chcon -t foo_exec_t bar # foo_exec_t is not yet defined chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument # cat setundefined.te policy_module(setundefined, 1.0) require { type unconfined_t; type unlabeled_t; } files_type(unlabeled_t) allow unconfined_t self:capability2 mac_admin; # make -f /usr/share/selinux/devel/Makefile setundefined.pp # semodule -i setundefined.pp # chcon -t foo_exec_t bar # foo_exec_t is not yet defined # mkdir -Z system_u:object_r:foo_exec_t baz # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # cat foo.te policy_module(foo, 1.0) type foo_exec_t; files_type(foo_exec_t) # make -f /usr/share/selinux/devel/Makefile foo.pp # semodule -i foo.pp # defines foo_exec_t # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r foo # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # semodule -i foo.pp # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r setundefined foo # chcon -t foo_exec_t bar # no longer defined and not allowed chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # rmdir baz # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2008-05-08 01:03:20 +08:00
rc = security_context_to_sid_force(value, size, &newsid);
if (rc) {
selinux: support deferred mapping of contexts Introduce SELinux support for deferred mapping of security contexts in the SID table upon policy reload, and use this support for inode security contexts when the context is not yet valid under the current policy. Only processes with CAP_MAC_ADMIN + mac_admin permission in policy can set undefined security contexts on inodes. Inodes with such undefined contexts are treated as having the unlabeled context until the context becomes valid upon a policy reload that defines the context. Context invalidation upon policy reload also uses this support to save the context information in the SID table and later recover it upon a subsequent policy reload that defines the context again. This support is to enable package managers and similar programs to set down file contexts unknown to the system policy at the time the file is created in order to better support placing loadable policy modules in packages and to support build systems that need to create images of different distro releases with different policies w/o requiring all of the contexts to be defined or legal in the build host policy. With this patch applied, the following sequence is possible, although in practice it is recommended that this permission only be allowed to specific program domains such as the package manager. # rmdir baz # rm bar # touch bar # chcon -t foo_exec_t bar # foo_exec_t is not yet defined chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument # cat setundefined.te policy_module(setundefined, 1.0) require { type unconfined_t; type unlabeled_t; } files_type(unlabeled_t) allow unconfined_t self:capability2 mac_admin; # make -f /usr/share/selinux/devel/Makefile setundefined.pp # semodule -i setundefined.pp # chcon -t foo_exec_t bar # foo_exec_t is not yet defined # mkdir -Z system_u:object_r:foo_exec_t baz # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # cat foo.te policy_module(foo, 1.0) type foo_exec_t; files_type(foo_exec_t) # make -f /usr/share/selinux/devel/Makefile foo.pp # semodule -i foo.pp # defines foo_exec_t # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r foo # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # semodule -i foo.pp # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r setundefined foo # chcon -t foo_exec_t bar # no longer defined and not allowed chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # rmdir baz # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2008-05-08 01:03:20 +08:00
printk(KERN_ERR "SELinux: unable to map context to SID"
"for (%s, %lu), rc=%d\n",
inode->i_sb->s_id, inode->i_ino, -rc);
return;
}
isec->sid = newsid;
return;
}
static int selinux_inode_getxattr(struct dentry *dentry, const char *name)
{
return dentry_has_perm(current, NULL, dentry, FILE__GETATTR);
}
static int selinux_inode_listxattr(struct dentry *dentry)
{
return dentry_has_perm(current, NULL, dentry, FILE__GETATTR);
}
static int selinux_inode_removexattr(struct dentry *dentry, const char *name)
{
Implement file posix capabilities Implement file posix capabilities. This allows programs to be given a subset of root's powers regardless of who runs them, without having to use setuid and giving the binary all of root's powers. This version works with Kaigai Kohei's userspace tools, found at http://www.kaigai.gr.jp/index.php. For more information on how to use this patch, Chris Friedhoff has posted a nice page at http://www.friedhoff.org/fscaps.html. Changelog: Nov 27: Incorporate fixes from Andrew Morton (security-introduce-file-caps-tweaks and security-introduce-file-caps-warning-fix) Fix Kconfig dependency. Fix change signaling behavior when file caps are not compiled in. Nov 13: Integrate comments from Alexey: Remove CONFIG_ ifdef from capability.h, and use %zd for printing a size_t. Nov 13: Fix endianness warnings by sparse as suggested by Alexey Dobriyan. Nov 09: Address warnings of unused variables at cap_bprm_set_security when file capabilities are disabled, and simultaneously clean up the code a little, by pulling the new code into a helper function. Nov 08: For pointers to required userspace tools and how to use them, see http://www.friedhoff.org/fscaps.html. Nov 07: Fix the calculation of the highest bit checked in check_cap_sanity(). Nov 07: Allow file caps to be enabled without CONFIG_SECURITY, since capabilities are the default. Hook cap_task_setscheduler when !CONFIG_SECURITY. Move capable(TASK_KILL) to end of cap_task_kill to reduce audit messages. Nov 05: Add secondary calls in selinux/hooks.c to task_setioprio and task_setscheduler so that selinux and capabilities with file cap support can be stacked. Sep 05: As Seth Arnold points out, uid checks are out of place for capability code. Sep 01: Define task_setscheduler, task_setioprio, cap_task_kill, and task_setnice to make sure a user cannot affect a process in which they called a program with some fscaps. One remaining question is the note under task_setscheduler: are we ok with CAP_SYS_NICE being sufficient to confine a process to a cpuset? It is a semantic change, as without fsccaps, attach_task doesn't allow CAP_SYS_NICE to override the uid equivalence check. But since it uses security_task_setscheduler, which elsewhere is used where CAP_SYS_NICE can be used to override the uid equivalence check, fixing it might be tough. task_setscheduler note: this also controls cpuset:attach_task. Are we ok with CAP_SYS_NICE being used to confine to a cpuset? task_setioprio task_setnice sys_setpriority uses this (through set_one_prio) for another process. Need same checks as setrlimit Aug 21: Updated secureexec implementation to reflect the fact that euid and uid might be the same and nonzero, but the process might still have elevated caps. Aug 15: Handle endianness of xattrs. Enforce capability version match between kernel and disk. Enforce that no bits beyond the known max capability are set, else return -EPERM. With this extra processing, it may be worth reconsidering doing all the work at bprm_set_security rather than d_instantiate. Aug 10: Always call getxattr at bprm_set_security, rather than caching it at d_instantiate. [morgan@kernel.org: file-caps clean up for linux/capability.h] [bunk@kernel.org: unexport cap_inode_killpriv] Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Andrew Morgan <morgan@kernel.org> Signed-off-by: Andrew Morgan <morgan@kernel.org> Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 14:31:36 +08:00
if (strcmp(name, XATTR_NAME_SELINUX))
return selinux_inode_setotherxattr(dentry, name);
/* No one is allowed to remove a SELinux security label.
You can change the label, but all data must be labeled. */
return -EACCES;
}
/*
* Copy the inode security context value to the user.
*
* Permission check is handled by selinux_inode_getxattr hook.
*/
static int selinux_inode_getsecurity(const struct inode *inode, const char *name, void **buffer, bool alloc)
{
u32 size;
int error;
char *context = NULL;
struct task_security_struct *tsec = current->security;
struct inode_security_struct *isec = inode->i_security;
if (strcmp(name, XATTR_SELINUX_SUFFIX))
return -EOPNOTSUPP;
/*
* If the caller has CAP_MAC_ADMIN, then get the raw context
* value even if it is not defined by current policy; otherwise,
* use the in-core value under current policy.
* Use the non-auditing forms of the permission checks since
* getxattr may be called by unprivileged processes commonly
* and lack of permission just means that we fall back to the
* in-core context value, not a denial.
*/
error = secondary_ops->capable(current, CAP_MAC_ADMIN);
if (!error)
error = avc_has_perm_noaudit(tsec->sid, tsec->sid,
SECCLASS_CAPABILITY2,
CAPABILITY2__MAC_ADMIN,
0,
NULL);
if (!error)
error = security_sid_to_context_force(isec->sid, &context,
&size);
else
error = security_sid_to_context(isec->sid, &context, &size);
if (error)
return error;
error = size;
if (alloc) {
*buffer = context;
goto out_nofree;
}
kfree(context);
out_nofree:
return error;
}
static int selinux_inode_setsecurity(struct inode *inode, const char *name,
const void *value, size_t size, int flags)
{
struct inode_security_struct *isec = inode->i_security;
u32 newsid;
int rc;
if (strcmp(name, XATTR_SELINUX_SUFFIX))
return -EOPNOTSUPP;
if (!value || !size)
return -EACCES;
rc = security_context_to_sid((void *)value, size, &newsid);
if (rc)
return rc;
isec->sid = newsid;
return 0;
}
static int selinux_inode_listsecurity(struct inode *inode, char *buffer, size_t buffer_size)
{
const int len = sizeof(XATTR_NAME_SELINUX);
if (buffer && len <= buffer_size)
memcpy(buffer, XATTR_NAME_SELINUX, len);
return len;
}
Implement file posix capabilities Implement file posix capabilities. This allows programs to be given a subset of root's powers regardless of who runs them, without having to use setuid and giving the binary all of root's powers. This version works with Kaigai Kohei's userspace tools, found at http://www.kaigai.gr.jp/index.php. For more information on how to use this patch, Chris Friedhoff has posted a nice page at http://www.friedhoff.org/fscaps.html. Changelog: Nov 27: Incorporate fixes from Andrew Morton (security-introduce-file-caps-tweaks and security-introduce-file-caps-warning-fix) Fix Kconfig dependency. Fix change signaling behavior when file caps are not compiled in. Nov 13: Integrate comments from Alexey: Remove CONFIG_ ifdef from capability.h, and use %zd for printing a size_t. Nov 13: Fix endianness warnings by sparse as suggested by Alexey Dobriyan. Nov 09: Address warnings of unused variables at cap_bprm_set_security when file capabilities are disabled, and simultaneously clean up the code a little, by pulling the new code into a helper function. Nov 08: For pointers to required userspace tools and how to use them, see http://www.friedhoff.org/fscaps.html. Nov 07: Fix the calculation of the highest bit checked in check_cap_sanity(). Nov 07: Allow file caps to be enabled without CONFIG_SECURITY, since capabilities are the default. Hook cap_task_setscheduler when !CONFIG_SECURITY. Move capable(TASK_KILL) to end of cap_task_kill to reduce audit messages. Nov 05: Add secondary calls in selinux/hooks.c to task_setioprio and task_setscheduler so that selinux and capabilities with file cap support can be stacked. Sep 05: As Seth Arnold points out, uid checks are out of place for capability code. Sep 01: Define task_setscheduler, task_setioprio, cap_task_kill, and task_setnice to make sure a user cannot affect a process in which they called a program with some fscaps. One remaining question is the note under task_setscheduler: are we ok with CAP_SYS_NICE being sufficient to confine a process to a cpuset? It is a semantic change, as without fsccaps, attach_task doesn't allow CAP_SYS_NICE to override the uid equivalence check. But since it uses security_task_setscheduler, which elsewhere is used where CAP_SYS_NICE can be used to override the uid equivalence check, fixing it might be tough. task_setscheduler note: this also controls cpuset:attach_task. Are we ok with CAP_SYS_NICE being used to confine to a cpuset? task_setioprio task_setnice sys_setpriority uses this (through set_one_prio) for another process. Need same checks as setrlimit Aug 21: Updated secureexec implementation to reflect the fact that euid and uid might be the same and nonzero, but the process might still have elevated caps. Aug 15: Handle endianness of xattrs. Enforce capability version match between kernel and disk. Enforce that no bits beyond the known max capability are set, else return -EPERM. With this extra processing, it may be worth reconsidering doing all the work at bprm_set_security rather than d_instantiate. Aug 10: Always call getxattr at bprm_set_security, rather than caching it at d_instantiate. [morgan@kernel.org: file-caps clean up for linux/capability.h] [bunk@kernel.org: unexport cap_inode_killpriv] Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Andrew Morgan <morgan@kernel.org> Signed-off-by: Andrew Morgan <morgan@kernel.org> Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 14:31:36 +08:00
static int selinux_inode_need_killpriv(struct dentry *dentry)
{
return secondary_ops->inode_need_killpriv(dentry);
}
static int selinux_inode_killpriv(struct dentry *dentry)
{
return secondary_ops->inode_killpriv(dentry);
}
static void selinux_inode_getsecid(const struct inode *inode, u32 *secid)
{
struct inode_security_struct *isec = inode->i_security;
*secid = isec->sid;
}
/* file security operations */
static int selinux_revalidate_file_permission(struct file *file, int mask)
{
int rc;
struct inode *inode = file->f_path.dentry->d_inode;
if (!mask) {
/* No permission to check. Existence test. */
return 0;
}
/* file_mask_to_av won't add FILE__WRITE if MAY_APPEND is set */
if ((file->f_flags & O_APPEND) && (mask & MAY_WRITE))
mask |= MAY_APPEND;
rc = file_has_perm(current, file,
file_mask_to_av(inode->i_mode, mask));
if (rc)
return rc;
return selinux_netlbl_inode_permission(inode, mask);
}
static int selinux_file_permission(struct file *file, int mask)
{
struct inode *inode = file->f_path.dentry->d_inode;
struct task_security_struct *tsec = current->security;
struct file_security_struct *fsec = file->f_security;
struct inode_security_struct *isec = inode->i_security;
if (!mask) {
/* No permission to check. Existence test. */
return 0;
}
if (tsec->sid == fsec->sid && fsec->isid == isec->sid
&& fsec->pseqno == avc_policy_seqno())
return selinux_netlbl_inode_permission(inode, mask);
return selinux_revalidate_file_permission(file, mask);
}
static int selinux_file_alloc_security(struct file *file)
{
return file_alloc_security(file);
}
static void selinux_file_free_security(struct file *file)
{
file_free_security(file);
}
static int selinux_file_ioctl(struct file *file, unsigned int cmd,
unsigned long arg)
{
u32 av = 0;
if (_IOC_DIR(cmd) & _IOC_WRITE)
av |= FILE__WRITE;
if (_IOC_DIR(cmd) & _IOC_READ)
av |= FILE__READ;
if (!av)
av = FILE__IOCTL;
return file_has_perm(current, file, av);
}
static int file_map_prot_check(struct file *file, unsigned long prot, int shared)
{
#ifndef CONFIG_PPC32
if ((prot & PROT_EXEC) && (!file || (!shared && (prot & PROT_WRITE)))) {
/*
* We are making executable an anonymous mapping or a
* private file mapping that will also be writable.
* This has an additional check.
*/
int rc = task_has_perm(current, current, PROCESS__EXECMEM);
if (rc)
return rc;
}
#endif
if (file) {
/* read access is always possible with a mapping */
u32 av = FILE__READ;
/* write access only matters if the mapping is shared */
if (shared && (prot & PROT_WRITE))
av |= FILE__WRITE;
if (prot & PROT_EXEC)
av |= FILE__EXECUTE;
return file_has_perm(current, file, av);
}
return 0;
}
static int selinux_file_mmap(struct file *file, unsigned long reqprot,
unsigned long prot, unsigned long flags,
unsigned long addr, unsigned long addr_only)
{
int rc = 0;
u32 sid = ((struct task_security_struct *)(current->security))->sid;
if (addr < mmap_min_addr)
rc = avc_has_perm(sid, sid, SECCLASS_MEMPROTECT,
MEMPROTECT__MMAP_ZERO, NULL);
if (rc || addr_only)
return rc;
if (selinux_checkreqprot)
prot = reqprot;
return file_map_prot_check(file, prot,
(flags & MAP_TYPE) == MAP_SHARED);
}
static int selinux_file_mprotect(struct vm_area_struct *vma,
unsigned long reqprot,
unsigned long prot)
{
int rc;
rc = secondary_ops->file_mprotect(vma, reqprot, prot);
if (rc)
return rc;
if (selinux_checkreqprot)
prot = reqprot;
#ifndef CONFIG_PPC32
if ((prot & PROT_EXEC) && !(vma->vm_flags & VM_EXEC)) {
rc = 0;
if (vma->vm_start >= vma->vm_mm->start_brk &&
vma->vm_end <= vma->vm_mm->brk) {
rc = task_has_perm(current, current,
PROCESS__EXECHEAP);
} else if (!vma->vm_file &&
vma->vm_start <= vma->vm_mm->start_stack &&
vma->vm_end >= vma->vm_mm->start_stack) {
rc = task_has_perm(current, current, PROCESS__EXECSTACK);
} else if (vma->vm_file && vma->anon_vma) {
/*
* We are making executable a file mapping that has
* had some COW done. Since pages might have been
* written, check ability to execute the possibly
* modified content. This typically should only
* occur for text relocations.
*/
rc = file_has_perm(current, vma->vm_file,
FILE__EXECMOD);
}
[PATCH] selinux: add executable stack check This patch adds an execstack permission check that controls the ability to make the main process stack executable so that attempts to make the stack executable can still be prevented even if the process is allowed the existing execmem permission in order to e.g. perform runtime code generation. Note that this does not yet address thread stacks. Note also that unlike the execmem check, the execstack check is only applied on mprotect calls, not mmap calls, as the current security_file_mmap hook is not passed the necessary information presently. The original author of the code that makes the distinction of the stack region, is Ingo Molnar, who wrote it within his patch for /proc/<pid>/maps markers. (http://marc.theaimsgroup.com/?l=linux-kernel&m=110719881508591&w=2) The patches also can be found at: http://pearls.tuxedo-es.org/patches/selinux/policy-execstack.patch http://pearls.tuxedo-es.org/patches/selinux/kernel-execstack.patch policy-execstack.patch is the patch that needs to be applied to the policy in order to support the execstack permission and exclude it from general_domain_access within macros/core_macros.te. kernel-execstack.patch adds such permission to the SELinux code within the kernel and adds the proper permission check to the selinux_file_mprotect() hook. Signed-off-by: Lorenzo Hernandez Garcia-Hierro <lorenzo@gnu.org> Acked-by: James Morris <jmorris@redhat.com> Acked-by: Stephen Smalley <sds@tycho.nsa.gov> Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-26 05:54:34 +08:00
if (rc)
return rc;
}
#endif
return file_map_prot_check(vma->vm_file, prot, vma->vm_flags&VM_SHARED);
}
static int selinux_file_lock(struct file *file, unsigned int cmd)
{
return file_has_perm(current, file, FILE__LOCK);
}
static int selinux_file_fcntl(struct file *file, unsigned int cmd,
unsigned long arg)
{
int err = 0;
switch (cmd) {
case F_SETFL:
if (!file->f_path.dentry || !file->f_path.dentry->d_inode) {
err = -EINVAL;
break;
}
if ((file->f_flags & O_APPEND) && !(arg & O_APPEND)) {
err = file_has_perm(current, file, FILE__WRITE);
break;
}
/* fall through */
case F_SETOWN:
case F_SETSIG:
case F_GETFL:
case F_GETOWN:
case F_GETSIG:
/* Just check FD__USE permission */
err = file_has_perm(current, file, 0);
break;
case F_GETLK:
case F_SETLK:
case F_SETLKW:
#if BITS_PER_LONG == 32
case F_GETLK64:
case F_SETLK64:
case F_SETLKW64:
#endif
if (!file->f_path.dentry || !file->f_path.dentry->d_inode) {
err = -EINVAL;
break;
}
err = file_has_perm(current, file, FILE__LOCK);
break;
}
return err;
}
static int selinux_file_set_fowner(struct file *file)
{
struct task_security_struct *tsec;
struct file_security_struct *fsec;
tsec = current->security;
fsec = file->f_security;
fsec->fown_sid = tsec->sid;
return 0;
}
static int selinux_file_send_sigiotask(struct task_struct *tsk,
struct fown_struct *fown, int signum)
{
struct file *file;
u32 perm;
struct task_security_struct *tsec;
struct file_security_struct *fsec;
/* struct fown_struct is never outside the context of a struct file */
file = container_of(fown, struct file, f_owner);
tsec = tsk->security;
fsec = file->f_security;
if (!signum)
perm = signal_to_av(SIGIO); /* as per send_sigio_to_task */
else
perm = signal_to_av(signum);
return avc_has_perm(fsec->fown_sid, tsec->sid,
SECCLASS_PROCESS, perm, NULL);
}
static int selinux_file_receive(struct file *file)
{
return file_has_perm(current, file, file_to_av(file));
}
static int selinux_dentry_open(struct file *file)
{
struct file_security_struct *fsec;
struct inode *inode;
struct inode_security_struct *isec;
inode = file->f_path.dentry->d_inode;
fsec = file->f_security;
isec = inode->i_security;
/*
* Save inode label and policy sequence number
* at open-time so that selinux_file_permission
* can determine whether revalidation is necessary.
* Task label is already saved in the file security
* struct as its SID.
*/
fsec->isid = isec->sid;
fsec->pseqno = avc_policy_seqno();
/*
* Since the inode label or policy seqno may have changed
* between the selinux_inode_permission check and the saving
* of state above, recheck that access is still permitted.
* Otherwise, access might never be revalidated against the
* new inode label or new policy.
* This check is not redundant - do not remove.
*/
return inode_has_perm(current, inode, file_to_av(file), NULL);
}
/* task security operations */
static int selinux_task_create(unsigned long clone_flags)
{
int rc;
rc = secondary_ops->task_create(clone_flags);
if (rc)
return rc;
return task_has_perm(current, current, PROCESS__FORK);
}
static int selinux_task_alloc_security(struct task_struct *tsk)
{
struct task_security_struct *tsec1, *tsec2;
int rc;
tsec1 = current->security;
rc = task_alloc_security(tsk);
if (rc)
return rc;
tsec2 = tsk->security;
tsec2->osid = tsec1->osid;
tsec2->sid = tsec1->sid;
/* Retain the exec, fs, key, and sock SIDs across fork */
tsec2->exec_sid = tsec1->exec_sid;
tsec2->create_sid = tsec1->create_sid;
tsec2->keycreate_sid = tsec1->keycreate_sid;
tsec2->sockcreate_sid = tsec1->sockcreate_sid;
return 0;
}
static void selinux_task_free_security(struct task_struct *tsk)
{
task_free_security(tsk);
}
static int selinux_task_setuid(uid_t id0, uid_t id1, uid_t id2, int flags)
{
/* Since setuid only affects the current process, and
since the SELinux controls are not based on the Linux
identity attributes, SELinux does not need to control
this operation. However, SELinux does control the use
of the CAP_SETUID and CAP_SETGID capabilities using the
capable hook. */
return 0;
}
static int selinux_task_post_setuid(uid_t id0, uid_t id1, uid_t id2, int flags)
{
return secondary_ops->task_post_setuid(id0, id1, id2, flags);
}
static int selinux_task_setgid(gid_t id0, gid_t id1, gid_t id2, int flags)
{
/* See the comment for setuid above. */
return 0;
}
static int selinux_task_setpgid(struct task_struct *p, pid_t pgid)
{
return task_has_perm(current, p, PROCESS__SETPGID);
}
static int selinux_task_getpgid(struct task_struct *p)
{
return task_has_perm(current, p, PROCESS__GETPGID);
}
static int selinux_task_getsid(struct task_struct *p)
{
return task_has_perm(current, p, PROCESS__GETSESSION);
}
static void selinux_task_getsecid(struct task_struct *p, u32 *secid)
{
struct task_security_struct *tsec = p->security;
*secid = tsec->sid;
}
static int selinux_task_setgroups(struct group_info *group_info)
{
/* See the comment for setuid above. */
return 0;
}
static int selinux_task_setnice(struct task_struct *p, int nice)
{
int rc;
rc = secondary_ops->task_setnice(p, nice);
if (rc)
return rc;
return task_has_perm(current, p, PROCESS__SETSCHED);
}
static int selinux_task_setioprio(struct task_struct *p, int ioprio)
{
Implement file posix capabilities Implement file posix capabilities. This allows programs to be given a subset of root's powers regardless of who runs them, without having to use setuid and giving the binary all of root's powers. This version works with Kaigai Kohei's userspace tools, found at http://www.kaigai.gr.jp/index.php. For more information on how to use this patch, Chris Friedhoff has posted a nice page at http://www.friedhoff.org/fscaps.html. Changelog: Nov 27: Incorporate fixes from Andrew Morton (security-introduce-file-caps-tweaks and security-introduce-file-caps-warning-fix) Fix Kconfig dependency. Fix change signaling behavior when file caps are not compiled in. Nov 13: Integrate comments from Alexey: Remove CONFIG_ ifdef from capability.h, and use %zd for printing a size_t. Nov 13: Fix endianness warnings by sparse as suggested by Alexey Dobriyan. Nov 09: Address warnings of unused variables at cap_bprm_set_security when file capabilities are disabled, and simultaneously clean up the code a little, by pulling the new code into a helper function. Nov 08: For pointers to required userspace tools and how to use them, see http://www.friedhoff.org/fscaps.html. Nov 07: Fix the calculation of the highest bit checked in check_cap_sanity(). Nov 07: Allow file caps to be enabled without CONFIG_SECURITY, since capabilities are the default. Hook cap_task_setscheduler when !CONFIG_SECURITY. Move capable(TASK_KILL) to end of cap_task_kill to reduce audit messages. Nov 05: Add secondary calls in selinux/hooks.c to task_setioprio and task_setscheduler so that selinux and capabilities with file cap support can be stacked. Sep 05: As Seth Arnold points out, uid checks are out of place for capability code. Sep 01: Define task_setscheduler, task_setioprio, cap_task_kill, and task_setnice to make sure a user cannot affect a process in which they called a program with some fscaps. One remaining question is the note under task_setscheduler: are we ok with CAP_SYS_NICE being sufficient to confine a process to a cpuset? It is a semantic change, as without fsccaps, attach_task doesn't allow CAP_SYS_NICE to override the uid equivalence check. But since it uses security_task_setscheduler, which elsewhere is used where CAP_SYS_NICE can be used to override the uid equivalence check, fixing it might be tough. task_setscheduler note: this also controls cpuset:attach_task. Are we ok with CAP_SYS_NICE being used to confine to a cpuset? task_setioprio task_setnice sys_setpriority uses this (through set_one_prio) for another process. Need same checks as setrlimit Aug 21: Updated secureexec implementation to reflect the fact that euid and uid might be the same and nonzero, but the process might still have elevated caps. Aug 15: Handle endianness of xattrs. Enforce capability version match between kernel and disk. Enforce that no bits beyond the known max capability are set, else return -EPERM. With this extra processing, it may be worth reconsidering doing all the work at bprm_set_security rather than d_instantiate. Aug 10: Always call getxattr at bprm_set_security, rather than caching it at d_instantiate. [morgan@kernel.org: file-caps clean up for linux/capability.h] [bunk@kernel.org: unexport cap_inode_killpriv] Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Andrew Morgan <morgan@kernel.org> Signed-off-by: Andrew Morgan <morgan@kernel.org> Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 14:31:36 +08:00
int rc;
rc = secondary_ops->task_setioprio(p, ioprio);
if (rc)
return rc;
return task_has_perm(current, p, PROCESS__SETSCHED);
}
static int selinux_task_getioprio(struct task_struct *p)
{
return task_has_perm(current, p, PROCESS__GETSCHED);
}
static int selinux_task_setrlimit(unsigned int resource, struct rlimit *new_rlim)
{
struct rlimit *old_rlim = current->signal->rlim + resource;
int rc;
rc = secondary_ops->task_setrlimit(resource, new_rlim);
if (rc)
return rc;
/* Control the ability to change the hard limit (whether
lowering or raising it), so that the hard limit can
later be used as a safe reset point for the soft limit
upon context transitions. See selinux_bprm_apply_creds. */
if (old_rlim->rlim_max != new_rlim->rlim_max)
return task_has_perm(current, current, PROCESS__SETRLIMIT);
return 0;
}
static int selinux_task_setscheduler(struct task_struct *p, int policy, struct sched_param *lp)
{
Implement file posix capabilities Implement file posix capabilities. This allows programs to be given a subset of root's powers regardless of who runs them, without having to use setuid and giving the binary all of root's powers. This version works with Kaigai Kohei's userspace tools, found at http://www.kaigai.gr.jp/index.php. For more information on how to use this patch, Chris Friedhoff has posted a nice page at http://www.friedhoff.org/fscaps.html. Changelog: Nov 27: Incorporate fixes from Andrew Morton (security-introduce-file-caps-tweaks and security-introduce-file-caps-warning-fix) Fix Kconfig dependency. Fix change signaling behavior when file caps are not compiled in. Nov 13: Integrate comments from Alexey: Remove CONFIG_ ifdef from capability.h, and use %zd for printing a size_t. Nov 13: Fix endianness warnings by sparse as suggested by Alexey Dobriyan. Nov 09: Address warnings of unused variables at cap_bprm_set_security when file capabilities are disabled, and simultaneously clean up the code a little, by pulling the new code into a helper function. Nov 08: For pointers to required userspace tools and how to use them, see http://www.friedhoff.org/fscaps.html. Nov 07: Fix the calculation of the highest bit checked in check_cap_sanity(). Nov 07: Allow file caps to be enabled without CONFIG_SECURITY, since capabilities are the default. Hook cap_task_setscheduler when !CONFIG_SECURITY. Move capable(TASK_KILL) to end of cap_task_kill to reduce audit messages. Nov 05: Add secondary calls in selinux/hooks.c to task_setioprio and task_setscheduler so that selinux and capabilities with file cap support can be stacked. Sep 05: As Seth Arnold points out, uid checks are out of place for capability code. Sep 01: Define task_setscheduler, task_setioprio, cap_task_kill, and task_setnice to make sure a user cannot affect a process in which they called a program with some fscaps. One remaining question is the note under task_setscheduler: are we ok with CAP_SYS_NICE being sufficient to confine a process to a cpuset? It is a semantic change, as without fsccaps, attach_task doesn't allow CAP_SYS_NICE to override the uid equivalence check. But since it uses security_task_setscheduler, which elsewhere is used where CAP_SYS_NICE can be used to override the uid equivalence check, fixing it might be tough. task_setscheduler note: this also controls cpuset:attach_task. Are we ok with CAP_SYS_NICE being used to confine to a cpuset? task_setioprio task_setnice sys_setpriority uses this (through set_one_prio) for another process. Need same checks as setrlimit Aug 21: Updated secureexec implementation to reflect the fact that euid and uid might be the same and nonzero, but the process might still have elevated caps. Aug 15: Handle endianness of xattrs. Enforce capability version match between kernel and disk. Enforce that no bits beyond the known max capability are set, else return -EPERM. With this extra processing, it may be worth reconsidering doing all the work at bprm_set_security rather than d_instantiate. Aug 10: Always call getxattr at bprm_set_security, rather than caching it at d_instantiate. [morgan@kernel.org: file-caps clean up for linux/capability.h] [bunk@kernel.org: unexport cap_inode_killpriv] Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Andrew Morgan <morgan@kernel.org> Signed-off-by: Andrew Morgan <morgan@kernel.org> Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 14:31:36 +08:00
int rc;
rc = secondary_ops->task_setscheduler(p, policy, lp);
if (rc)
return rc;
return task_has_perm(current, p, PROCESS__SETSCHED);
}
static int selinux_task_getscheduler(struct task_struct *p)
{
return task_has_perm(current, p, PROCESS__GETSCHED);
}
static int selinux_task_movememory(struct task_struct *p)
{
return task_has_perm(current, p, PROCESS__SETSCHED);
}
static int selinux_task_kill(struct task_struct *p, struct siginfo *info,
int sig, u32 secid)
{
u32 perm;
int rc;
struct task_security_struct *tsec;
rc = secondary_ops->task_kill(p, info, sig, secid);
if (rc)
return rc;
if (!sig)
perm = PROCESS__SIGNULL; /* null signal; existence test */
else
perm = signal_to_av(sig);
tsec = p->security;
if (secid)
rc = avc_has_perm(secid, tsec->sid, SECCLASS_PROCESS, perm, NULL);
else
rc = task_has_perm(current, p, perm);
return rc;
}
static int selinux_task_prctl(int option,
unsigned long arg2,
unsigned long arg3,
unsigned long arg4,
capabilities: implement per-process securebits Filesystem capability support makes it possible to do away with (set)uid-0 based privilege and use capabilities instead. That is, with filesystem support for capabilities but without this present patch, it is (conceptually) possible to manage a system with capabilities alone and never need to obtain privilege via (set)uid-0. Of course, conceptually isn't quite the same as currently possible since few user applications, certainly not enough to run a viable system, are currently prepared to leverage capabilities to exercise privilege. Further, many applications exist that may never get upgraded in this way, and the kernel will continue to want to support their setuid-0 base privilege needs. Where pure-capability applications evolve and replace setuid-0 binaries, it is desirable that there be a mechanisms by which they can contain their privilege. In addition to leveraging the per-process bounding and inheritable sets, this should include suppressing the privilege of the uid-0 superuser from the process' tree of children. The feature added by this patch can be leveraged to suppress the privilege associated with (set)uid-0. This suppression requires CAP_SETPCAP to initiate, and only immediately affects the 'current' process (it is inherited through fork()/exec()). This reimplementation differs significantly from the historical support for securebits which was system-wide, unwieldy and which has ultimately withered to a dead relic in the source of the modern kernel. With this patch applied a process, that is capable(CAP_SETPCAP), can now drop all legacy privilege (through uid=0) for itself and all subsequently fork()'d/exec()'d children with: prctl(PR_SET_SECUREBITS, 0x2f); This patch represents a no-op unless CONFIG_SECURITY_FILE_CAPABILITIES is enabled at configure time. [akpm@linux-foundation.org: fix uninitialised var warning] [serue@us.ibm.com: capabilities: use cap_task_prctl when !CONFIG_SECURITY] Signed-off-by: Andrew G. Morgan <morgan@kernel.org> Acked-by: Serge Hallyn <serue@us.ibm.com> Reviewed-by: James Morris <jmorris@namei.org> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: Paul Moore <paul.moore@hp.com> Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:40 +08:00
unsigned long arg5,
long *rc_p)
{
/* The current prctl operations do not appear to require
any SELinux controls since they merely observe or modify
the state of the current process. */
capabilities: implement per-process securebits Filesystem capability support makes it possible to do away with (set)uid-0 based privilege and use capabilities instead. That is, with filesystem support for capabilities but without this present patch, it is (conceptually) possible to manage a system with capabilities alone and never need to obtain privilege via (set)uid-0. Of course, conceptually isn't quite the same as currently possible since few user applications, certainly not enough to run a viable system, are currently prepared to leverage capabilities to exercise privilege. Further, many applications exist that may never get upgraded in this way, and the kernel will continue to want to support their setuid-0 base privilege needs. Where pure-capability applications evolve and replace setuid-0 binaries, it is desirable that there be a mechanisms by which they can contain their privilege. In addition to leveraging the per-process bounding and inheritable sets, this should include suppressing the privilege of the uid-0 superuser from the process' tree of children. The feature added by this patch can be leveraged to suppress the privilege associated with (set)uid-0. This suppression requires CAP_SETPCAP to initiate, and only immediately affects the 'current' process (it is inherited through fork()/exec()). This reimplementation differs significantly from the historical support for securebits which was system-wide, unwieldy and which has ultimately withered to a dead relic in the source of the modern kernel. With this patch applied a process, that is capable(CAP_SETPCAP), can now drop all legacy privilege (through uid=0) for itself and all subsequently fork()'d/exec()'d children with: prctl(PR_SET_SECUREBITS, 0x2f); This patch represents a no-op unless CONFIG_SECURITY_FILE_CAPABILITIES is enabled at configure time. [akpm@linux-foundation.org: fix uninitialised var warning] [serue@us.ibm.com: capabilities: use cap_task_prctl when !CONFIG_SECURITY] Signed-off-by: Andrew G. Morgan <morgan@kernel.org> Acked-by: Serge Hallyn <serue@us.ibm.com> Reviewed-by: James Morris <jmorris@namei.org> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: Paul Moore <paul.moore@hp.com> Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:40 +08:00
return secondary_ops->task_prctl(option, arg2, arg3, arg4, arg5, rc_p);
}
static int selinux_task_wait(struct task_struct *p)
{
return task_has_perm(p, current, PROCESS__SIGCHLD);
}
static void selinux_task_reparent_to_init(struct task_struct *p)
{
struct task_security_struct *tsec;
secondary_ops->task_reparent_to_init(p);
tsec = p->security;
tsec->osid = tsec->sid;
tsec->sid = SECINITSID_KERNEL;
return;
}
static void selinux_task_to_inode(struct task_struct *p,
struct inode *inode)
{
struct task_security_struct *tsec = p->security;
struct inode_security_struct *isec = inode->i_security;
isec->sid = tsec->sid;
isec->initialized = 1;
return;
}
/* Returns error only if unable to parse addresses */
static int selinux_parse_skb_ipv4(struct sk_buff *skb,
struct avc_audit_data *ad, u8 *proto)
{
int offset, ihlen, ret = -EINVAL;
struct iphdr _iph, *ih;
offset = skb_network_offset(skb);
ih = skb_header_pointer(skb, offset, sizeof(_iph), &_iph);
if (ih == NULL)
goto out;
ihlen = ih->ihl * 4;
if (ihlen < sizeof(_iph))
goto out;
ad->u.net.v4info.saddr = ih->saddr;
ad->u.net.v4info.daddr = ih->daddr;
ret = 0;
if (proto)
*proto = ih->protocol;
switch (ih->protocol) {
case IPPROTO_TCP: {
struct tcphdr _tcph, *th;
if (ntohs(ih->frag_off) & IP_OFFSET)
break;
offset += ihlen;
th = skb_header_pointer(skb, offset, sizeof(_tcph), &_tcph);
if (th == NULL)
break;
ad->u.net.sport = th->source;
ad->u.net.dport = th->dest;
break;
}
case IPPROTO_UDP: {
struct udphdr _udph, *uh;
if (ntohs(ih->frag_off) & IP_OFFSET)
break;
offset += ihlen;
uh = skb_header_pointer(skb, offset, sizeof(_udph), &_udph);
if (uh == NULL)
break;
ad->u.net.sport = uh->source;
ad->u.net.dport = uh->dest;
break;
}
case IPPROTO_DCCP: {
struct dccp_hdr _dccph, *dh;
if (ntohs(ih->frag_off) & IP_OFFSET)
break;
offset += ihlen;
dh = skb_header_pointer(skb, offset, sizeof(_dccph), &_dccph);
if (dh == NULL)
break;
ad->u.net.sport = dh->dccph_sport;
ad->u.net.dport = dh->dccph_dport;
break;
}
default:
break;
}
out:
return ret;
}
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
/* Returns error only if unable to parse addresses */
static int selinux_parse_skb_ipv6(struct sk_buff *skb,
struct avc_audit_data *ad, u8 *proto)
{
u8 nexthdr;
int ret = -EINVAL, offset;
struct ipv6hdr _ipv6h, *ip6;
offset = skb_network_offset(skb);
ip6 = skb_header_pointer(skb, offset, sizeof(_ipv6h), &_ipv6h);
if (ip6 == NULL)
goto out;
ipv6_addr_copy(&ad->u.net.v6info.saddr, &ip6->saddr);
ipv6_addr_copy(&ad->u.net.v6info.daddr, &ip6->daddr);
ret = 0;
nexthdr = ip6->nexthdr;
offset += sizeof(_ipv6h);
offset = ipv6_skip_exthdr(skb, offset, &nexthdr);
if (offset < 0)
goto out;
if (proto)
*proto = nexthdr;
switch (nexthdr) {
case IPPROTO_TCP: {
struct tcphdr _tcph, *th;
th = skb_header_pointer(skb, offset, sizeof(_tcph), &_tcph);
if (th == NULL)
break;
ad->u.net.sport = th->source;
ad->u.net.dport = th->dest;
break;
}
case IPPROTO_UDP: {
struct udphdr _udph, *uh;
uh = skb_header_pointer(skb, offset, sizeof(_udph), &_udph);
if (uh == NULL)
break;
ad->u.net.sport = uh->source;
ad->u.net.dport = uh->dest;
break;
}
case IPPROTO_DCCP: {
struct dccp_hdr _dccph, *dh;
dh = skb_header_pointer(skb, offset, sizeof(_dccph), &_dccph);
if (dh == NULL)
break;
ad->u.net.sport = dh->dccph_sport;
ad->u.net.dport = dh->dccph_dport;
break;
}
/* includes fragments */
default:
break;
}
out:
return ret;
}
#endif /* IPV6 */
static int selinux_parse_skb(struct sk_buff *skb, struct avc_audit_data *ad,
char **_addrp, int src, u8 *proto)
{
char *addrp;
int ret;
switch (ad->u.net.family) {
case PF_INET:
ret = selinux_parse_skb_ipv4(skb, ad, proto);
if (ret)
goto parse_error;
addrp = (char *)(src ? &ad->u.net.v4info.saddr :
&ad->u.net.v4info.daddr);
goto okay;
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
case PF_INET6:
ret = selinux_parse_skb_ipv6(skb, ad, proto);
if (ret)
goto parse_error;
addrp = (char *)(src ? &ad->u.net.v6info.saddr :
&ad->u.net.v6info.daddr);
goto okay;
#endif /* IPV6 */
default:
addrp = NULL;
goto okay;
}
parse_error:
printk(KERN_WARNING
"SELinux: failure in selinux_parse_skb(),"
" unable to parse packet\n");
return ret;
okay:
if (_addrp)
*_addrp = addrp;
return 0;
}
/**
* selinux_skb_peerlbl_sid - Determine the peer label of a packet
* @skb: the packet
* @family: protocol family
* @sid: the packet's peer label SID
*
* Description:
* Check the various different forms of network peer labeling and determine
* the peer label/SID for the packet; most of the magic actually occurs in
* the security server function security_net_peersid_cmp(). The function
* returns zero if the value in @sid is valid (although it may be SECSID_NULL)
* or -EACCES if @sid is invalid due to inconsistencies with the different
* peer labels.
*
*/
static int selinux_skb_peerlbl_sid(struct sk_buff *skb, u16 family, u32 *sid)
{
int err;
u32 xfrm_sid;
u32 nlbl_sid;
u32 nlbl_type;
selinux_skb_xfrm_sid(skb, &xfrm_sid);
selinux_netlbl_skbuff_getsid(skb, family, &nlbl_type, &nlbl_sid);
err = security_net_peersid_resolve(nlbl_sid, nlbl_type, xfrm_sid, sid);
if (unlikely(err)) {
printk(KERN_WARNING
"SELinux: failure in selinux_skb_peerlbl_sid(),"
" unable to determine packet's peer label\n");
return -EACCES;
}
return 0;
}
/* socket security operations */
static int socket_has_perm(struct task_struct *task, struct socket *sock,
u32 perms)
{
struct inode_security_struct *isec;
struct task_security_struct *tsec;
struct avc_audit_data ad;
int err = 0;
tsec = task->security;
isec = SOCK_INODE(sock)->i_security;
if (isec->sid == SECINITSID_KERNEL)
goto out;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.sk = sock->sk;
err = avc_has_perm(tsec->sid, isec->sid, isec->sclass, perms, &ad);
out:
return err;
}
static int selinux_socket_create(int family, int type,
int protocol, int kern)
{
int err = 0;
struct task_security_struct *tsec;
u32 newsid;
if (kern)
goto out;
tsec = current->security;
newsid = tsec->sockcreate_sid ? : tsec->sid;
err = avc_has_perm(tsec->sid, newsid,
socket_type_to_security_class(family, type,
protocol), SOCKET__CREATE, NULL);
out:
return err;
}
static int selinux_socket_post_create(struct socket *sock, int family,
int type, int protocol, int kern)
{
int err = 0;
struct inode_security_struct *isec;
struct task_security_struct *tsec;
struct sk_security_struct *sksec;
u32 newsid;
isec = SOCK_INODE(sock)->i_security;
tsec = current->security;
newsid = tsec->sockcreate_sid ? : tsec->sid;
isec->sclass = socket_type_to_security_class(family, type, protocol);
isec->sid = kern ? SECINITSID_KERNEL : newsid;
isec->initialized = 1;
if (sock->sk) {
sksec = sock->sk->sk_security;
sksec->sid = isec->sid;
sksec->sclass = isec->sclass;
err = selinux_netlbl_socket_post_create(sock);
}
return err;
}
/* Range of port numbers used to automatically bind.
Need to determine whether we should perform a name_bind
permission check between the socket and the port number. */
static int selinux_socket_bind(struct socket *sock, struct sockaddr *address, int addrlen)
{
u16 family;
int err;
err = socket_has_perm(current, sock, SOCKET__BIND);
if (err)
goto out;
/*
* If PF_INET or PF_INET6, check name_bind permission for the port.
[PATCH] SELinux - fix SCTP socket bug and general IP protocol handling The following patch updates the way SELinux classifies and handles IP based protocols. Currently, IP sockets are classified by SELinux as being either TCP, UDP or 'Raw', the latter being a default for IP socket that is not TCP or UDP. The classification code is out of date and uses only the socket type parameter to socket(2) to determine the class of IP socket. So, any socket created with SOCK_STREAM will be classified by SELinux as TCP, and SOCK_DGRAM as UDP. Also, other socket types such as SOCK_SEQPACKET and SOCK_DCCP are currently ignored by SELinux, which classifies them as generic sockets, which means they don't even get basic IP level checking. This patch changes the SELinux IP socket classification logic, so that only an IPPROTO_IP protocol value passed to socket(2) classify the socket as TCP or UDP. The patch also drops the check for SOCK_RAW and converts it into a default, so that socket types like SOCK_DCCP and SOCK_SEQPACKET are classified as SECCLASS_RAWIP_SOCKET (instead of generic sockets). Note that protocol-specific support for SCTP, DCCP etc. is not addressed here, we're just getting these protocols checked at the IP layer. This fixes a reported problem where SCTP sockets were being recognized as generic SELinux sockets yet still being passed in one case to an IP level check, which then fails for generic sockets. It will also fix bugs where any SOCK_STREAM socket is classified as TCP or any SOCK_DGRAM socket is classified as UDP. This patch also unifies the way IP sockets classes are determined in selinux_socket_bind(), so we use the already calculated value instead of trying to recalculate it. Signed-off-by: James Morris <jmorris@namei.org> Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-01 02:24:34 +08:00
* Multiple address binding for SCTP is not supported yet: we just
* check the first address now.
*/
family = sock->sk->sk_family;
if (family == PF_INET || family == PF_INET6) {
char *addrp;
struct inode_security_struct *isec;
struct task_security_struct *tsec;
struct avc_audit_data ad;
struct sockaddr_in *addr4 = NULL;
struct sockaddr_in6 *addr6 = NULL;
unsigned short snum;
struct sock *sk = sock->sk;
u32 sid, node_perm;
tsec = current->security;
isec = SOCK_INODE(sock)->i_security;
if (family == PF_INET) {
addr4 = (struct sockaddr_in *)address;
snum = ntohs(addr4->sin_port);
addrp = (char *)&addr4->sin_addr.s_addr;
} else {
addr6 = (struct sockaddr_in6 *)address;
snum = ntohs(addr6->sin6_port);
addrp = (char *)&addr6->sin6_addr.s6_addr;
}
if (snum) {
int low, high;
inet_get_local_port_range(&low, &high);
if (snum < max(PROT_SOCK, low) || snum > high) {
err = sel_netport_sid(sk->sk_protocol,
snum, &sid);
if (err)
goto out;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.sport = htons(snum);
ad.u.net.family = family;
err = avc_has_perm(isec->sid, sid,
isec->sclass,
SOCKET__NAME_BIND, &ad);
if (err)
goto out;
}
}
switch (isec->sclass) {
[PATCH] SELinux - fix SCTP socket bug and general IP protocol handling The following patch updates the way SELinux classifies and handles IP based protocols. Currently, IP sockets are classified by SELinux as being either TCP, UDP or 'Raw', the latter being a default for IP socket that is not TCP or UDP. The classification code is out of date and uses only the socket type parameter to socket(2) to determine the class of IP socket. So, any socket created with SOCK_STREAM will be classified by SELinux as TCP, and SOCK_DGRAM as UDP. Also, other socket types such as SOCK_SEQPACKET and SOCK_DCCP are currently ignored by SELinux, which classifies them as generic sockets, which means they don't even get basic IP level checking. This patch changes the SELinux IP socket classification logic, so that only an IPPROTO_IP protocol value passed to socket(2) classify the socket as TCP or UDP. The patch also drops the check for SOCK_RAW and converts it into a default, so that socket types like SOCK_DCCP and SOCK_SEQPACKET are classified as SECCLASS_RAWIP_SOCKET (instead of generic sockets). Note that protocol-specific support for SCTP, DCCP etc. is not addressed here, we're just getting these protocols checked at the IP layer. This fixes a reported problem where SCTP sockets were being recognized as generic SELinux sockets yet still being passed in one case to an IP level check, which then fails for generic sockets. It will also fix bugs where any SOCK_STREAM socket is classified as TCP or any SOCK_DGRAM socket is classified as UDP. This patch also unifies the way IP sockets classes are determined in selinux_socket_bind(), so we use the already calculated value instead of trying to recalculate it. Signed-off-by: James Morris <jmorris@namei.org> Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-01 02:24:34 +08:00
case SECCLASS_TCP_SOCKET:
node_perm = TCP_SOCKET__NODE_BIND;
break;
[PATCH] SELinux - fix SCTP socket bug and general IP protocol handling The following patch updates the way SELinux classifies and handles IP based protocols. Currently, IP sockets are classified by SELinux as being either TCP, UDP or 'Raw', the latter being a default for IP socket that is not TCP or UDP. The classification code is out of date and uses only the socket type parameter to socket(2) to determine the class of IP socket. So, any socket created with SOCK_STREAM will be classified by SELinux as TCP, and SOCK_DGRAM as UDP. Also, other socket types such as SOCK_SEQPACKET and SOCK_DCCP are currently ignored by SELinux, which classifies them as generic sockets, which means they don't even get basic IP level checking. This patch changes the SELinux IP socket classification logic, so that only an IPPROTO_IP protocol value passed to socket(2) classify the socket as TCP or UDP. The patch also drops the check for SOCK_RAW and converts it into a default, so that socket types like SOCK_DCCP and SOCK_SEQPACKET are classified as SECCLASS_RAWIP_SOCKET (instead of generic sockets). Note that protocol-specific support for SCTP, DCCP etc. is not addressed here, we're just getting these protocols checked at the IP layer. This fixes a reported problem where SCTP sockets were being recognized as generic SELinux sockets yet still being passed in one case to an IP level check, which then fails for generic sockets. It will also fix bugs where any SOCK_STREAM socket is classified as TCP or any SOCK_DGRAM socket is classified as UDP. This patch also unifies the way IP sockets classes are determined in selinux_socket_bind(), so we use the already calculated value instead of trying to recalculate it. Signed-off-by: James Morris <jmorris@namei.org> Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-01 02:24:34 +08:00
case SECCLASS_UDP_SOCKET:
node_perm = UDP_SOCKET__NODE_BIND;
break;
case SECCLASS_DCCP_SOCKET:
node_perm = DCCP_SOCKET__NODE_BIND;
break;
default:
node_perm = RAWIP_SOCKET__NODE_BIND;
break;
}
err = sel_netnode_sid(addrp, family, &sid);
if (err)
goto out;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.sport = htons(snum);
ad.u.net.family = family;
if (family == PF_INET)
ad.u.net.v4info.saddr = addr4->sin_addr.s_addr;
else
ipv6_addr_copy(&ad.u.net.v6info.saddr, &addr6->sin6_addr);
err = avc_has_perm(isec->sid, sid,
isec->sclass, node_perm, &ad);
if (err)
goto out;
}
out:
return err;
}
static int selinux_socket_connect(struct socket *sock, struct sockaddr *address, int addrlen)
{
struct sock *sk = sock->sk;
struct inode_security_struct *isec;
int err;
err = socket_has_perm(current, sock, SOCKET__CONNECT);
if (err)
return err;
/*
* If a TCP or DCCP socket, check name_connect permission for the port.
*/
isec = SOCK_INODE(sock)->i_security;
if (isec->sclass == SECCLASS_TCP_SOCKET ||
isec->sclass == SECCLASS_DCCP_SOCKET) {
struct avc_audit_data ad;
struct sockaddr_in *addr4 = NULL;
struct sockaddr_in6 *addr6 = NULL;
unsigned short snum;
u32 sid, perm;
if (sk->sk_family == PF_INET) {
addr4 = (struct sockaddr_in *)address;
if (addrlen < sizeof(struct sockaddr_in))
return -EINVAL;
snum = ntohs(addr4->sin_port);
} else {
addr6 = (struct sockaddr_in6 *)address;
if (addrlen < SIN6_LEN_RFC2133)
return -EINVAL;
snum = ntohs(addr6->sin6_port);
}
err = sel_netport_sid(sk->sk_protocol, snum, &sid);
if (err)
goto out;
perm = (isec->sclass == SECCLASS_TCP_SOCKET) ?
TCP_SOCKET__NAME_CONNECT : DCCP_SOCKET__NAME_CONNECT;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.dport = htons(snum);
ad.u.net.family = sk->sk_family;
err = avc_has_perm(isec->sid, sid, isec->sclass, perm, &ad);
if (err)
goto out;
}
err = selinux_netlbl_socket_connect(sk, address);
out:
return err;
}
static int selinux_socket_listen(struct socket *sock, int backlog)
{
return socket_has_perm(current, sock, SOCKET__LISTEN);
}
static int selinux_socket_accept(struct socket *sock, struct socket *newsock)
{
int err;
struct inode_security_struct *isec;
struct inode_security_struct *newisec;
err = socket_has_perm(current, sock, SOCKET__ACCEPT);
if (err)
return err;
newisec = SOCK_INODE(newsock)->i_security;
isec = SOCK_INODE(sock)->i_security;
newisec->sclass = isec->sclass;
newisec->sid = isec->sid;
newisec->initialized = 1;
return 0;
}
static int selinux_socket_sendmsg(struct socket *sock, struct msghdr *msg,
int size)
{
int rc;
rc = socket_has_perm(current, sock, SOCKET__WRITE);
if (rc)
return rc;
return selinux_netlbl_inode_permission(SOCK_INODE(sock), MAY_WRITE);
}
static int selinux_socket_recvmsg(struct socket *sock, struct msghdr *msg,
int size, int flags)
{
return socket_has_perm(current, sock, SOCKET__READ);
}
static int selinux_socket_getsockname(struct socket *sock)
{
return socket_has_perm(current, sock, SOCKET__GETATTR);
}
static int selinux_socket_getpeername(struct socket *sock)
{
return socket_has_perm(current, sock, SOCKET__GETATTR);
}
static int selinux_socket_setsockopt(struct socket *sock, int level, int optname)
{
int err;
err = socket_has_perm(current, sock, SOCKET__SETOPT);
if (err)
return err;
return selinux_netlbl_socket_setsockopt(sock, level, optname);
}
static int selinux_socket_getsockopt(struct socket *sock, int level,
int optname)
{
return socket_has_perm(current, sock, SOCKET__GETOPT);
}
static int selinux_socket_shutdown(struct socket *sock, int how)
{
return socket_has_perm(current, sock, SOCKET__SHUTDOWN);
}
static int selinux_socket_unix_stream_connect(struct socket *sock,
struct socket *other,
struct sock *newsk)
{
struct sk_security_struct *ssec;
struct inode_security_struct *isec;
struct inode_security_struct *other_isec;
struct avc_audit_data ad;
int err;
err = secondary_ops->unix_stream_connect(sock, other, newsk);
if (err)
return err;
isec = SOCK_INODE(sock)->i_security;
other_isec = SOCK_INODE(other)->i_security;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.sk = other->sk;
err = avc_has_perm(isec->sid, other_isec->sid,
isec->sclass,
UNIX_STREAM_SOCKET__CONNECTTO, &ad);
if (err)
return err;
/* connecting socket */
ssec = sock->sk->sk_security;
ssec->peer_sid = other_isec->sid;
/* server child socket */
ssec = newsk->sk_security;
ssec->peer_sid = isec->sid;
err = security_sid_mls_copy(other_isec->sid, ssec->peer_sid, &ssec->sid);
return err;
}
static int selinux_socket_unix_may_send(struct socket *sock,
struct socket *other)
{
struct inode_security_struct *isec;
struct inode_security_struct *other_isec;
struct avc_audit_data ad;
int err;
isec = SOCK_INODE(sock)->i_security;
other_isec = SOCK_INODE(other)->i_security;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.sk = other->sk;
err = avc_has_perm(isec->sid, other_isec->sid,
isec->sclass, SOCKET__SENDTO, &ad);
if (err)
return err;
return 0;
}
static int selinux_inet_sys_rcv_skb(int ifindex, char *addrp, u16 family,
u32 peer_sid,
struct avc_audit_data *ad)
{
int err;
u32 if_sid;
u32 node_sid;
err = sel_netif_sid(ifindex, &if_sid);
if (err)
return err;
err = avc_has_perm(peer_sid, if_sid,
SECCLASS_NETIF, NETIF__INGRESS, ad);
if (err)
return err;
err = sel_netnode_sid(addrp, family, &node_sid);
if (err)
return err;
return avc_has_perm(peer_sid, node_sid,
SECCLASS_NODE, NODE__RECVFROM, ad);
}
static int selinux_sock_rcv_skb_iptables_compat(struct sock *sk,
struct sk_buff *skb,
struct avc_audit_data *ad,
u16 family,
char *addrp)
{
int err;
struct sk_security_struct *sksec = sk->sk_security;
u16 sk_class;
u32 netif_perm, node_perm, recv_perm;
u32 port_sid, node_sid, if_sid, sk_sid;
sk_sid = sksec->sid;
sk_class = sksec->sclass;
switch (sk_class) {
case SECCLASS_UDP_SOCKET:
netif_perm = NETIF__UDP_RECV;
node_perm = NODE__UDP_RECV;
recv_perm = UDP_SOCKET__RECV_MSG;
break;
case SECCLASS_TCP_SOCKET:
netif_perm = NETIF__TCP_RECV;
node_perm = NODE__TCP_RECV;
recv_perm = TCP_SOCKET__RECV_MSG;
break;
case SECCLASS_DCCP_SOCKET:
netif_perm = NETIF__DCCP_RECV;
node_perm = NODE__DCCP_RECV;
recv_perm = DCCP_SOCKET__RECV_MSG;
break;
default:
netif_perm = NETIF__RAWIP_RECV;
node_perm = NODE__RAWIP_RECV;
recv_perm = 0;
break;
}
err = sel_netif_sid(skb->iif, &if_sid);
if (err)
return err;
err = avc_has_perm(sk_sid, if_sid, SECCLASS_NETIF, netif_perm, ad);
if (err)
return err;
err = sel_netnode_sid(addrp, family, &node_sid);
if (err)
return err;
err = avc_has_perm(sk_sid, node_sid, SECCLASS_NODE, node_perm, ad);
if (err)
return err;
if (!recv_perm)
return 0;
err = sel_netport_sid(sk->sk_protocol,
ntohs(ad->u.net.sport), &port_sid);
if (unlikely(err)) {
printk(KERN_WARNING
"SELinux: failure in"
" selinux_sock_rcv_skb_iptables_compat(),"
" network port label not found\n");
return err;
}
return avc_has_perm(sk_sid, port_sid, sk_class, recv_perm, ad);
}
static int selinux_sock_rcv_skb_compat(struct sock *sk, struct sk_buff *skb,
u16 family)
{
int err;
struct sk_security_struct *sksec = sk->sk_security;
u32 peer_sid;
u32 sk_sid = sksec->sid;
struct avc_audit_data ad;
char *addrp;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.netif = skb->iif;
ad.u.net.family = family;
err = selinux_parse_skb(skb, &ad, &addrp, 1, NULL);
if (err)
return err;
if (selinux_compat_net)
err = selinux_sock_rcv_skb_iptables_compat(sk, skb, &ad,
family, addrp);
else
err = avc_has_perm(sk_sid, skb->secmark, SECCLASS_PACKET,
PACKET__RECV, &ad);
if (err)
return err;
if (selinux_policycap_netpeer) {
err = selinux_skb_peerlbl_sid(skb, family, &peer_sid);
if (err)
return err;
err = avc_has_perm(sk_sid, peer_sid,
SECCLASS_PEER, PEER__RECV, &ad);
if (err)
selinux_netlbl_err(skb, err, 0);
} else {
err = selinux_netlbl_sock_rcv_skb(sksec, skb, family, &ad);
if (err)
return err;
err = selinux_xfrm_sock_rcv_skb(sksec->sid, skb, &ad);
}
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
return err;
}
static int selinux_socket_sock_rcv_skb(struct sock *sk, struct sk_buff *skb)
{
int err;
struct sk_security_struct *sksec = sk->sk_security;
u16 family = sk->sk_family;
u32 sk_sid = sksec->sid;
struct avc_audit_data ad;
char *addrp;
u8 secmark_active;
u8 peerlbl_active;
if (family != PF_INET && family != PF_INET6)
return 0;
/* Handle mapped IPv4 packets arriving via IPv6 sockets */
if (family == PF_INET6 && skb->protocol == htons(ETH_P_IP))
family = PF_INET;
/* If any sort of compatibility mode is enabled then handoff processing
* to the selinux_sock_rcv_skb_compat() function to deal with the
* special handling. We do this in an attempt to keep this function
* as fast and as clean as possible. */
if (selinux_compat_net || !selinux_policycap_netpeer)
return selinux_sock_rcv_skb_compat(sk, skb, family);
secmark_active = selinux_secmark_enabled();
peerlbl_active = netlbl_enabled() || selinux_xfrm_enabled();
if (!secmark_active && !peerlbl_active)
return 0;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.netif = skb->iif;
ad.u.net.family = family;
err = selinux_parse_skb(skb, &ad, &addrp, 1, NULL);
if (err)
return err;
if (peerlbl_active) {
u32 peer_sid;
err = selinux_skb_peerlbl_sid(skb, family, &peer_sid);
if (err)
return err;
err = selinux_inet_sys_rcv_skb(skb->iif, addrp, family,
peer_sid, &ad);
if (err) {
selinux_netlbl_err(skb, err, 0);
return err;
}
err = avc_has_perm(sk_sid, peer_sid, SECCLASS_PEER,
PEER__RECV, &ad);
if (err)
selinux_netlbl_err(skb, err, 0);
}
if (secmark_active) {
err = avc_has_perm(sk_sid, skb->secmark, SECCLASS_PACKET,
PACKET__RECV, &ad);
if (err)
return err;
}
return err;
}
[SECURITY]: TCP/UDP getpeersec This patch implements an application of the LSM-IPSec networking controls whereby an application can determine the label of the security association its TCP or UDP sockets are currently connected to via getsockopt and the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of an IPSec security association a particular TCP or UDP socket is using. The application can then use this security context to determine the security context for processing on behalf of the peer at the other end of this connection. In the case of UDP, the security context is for each individual packet. An example application is the inetd daemon, which could be modified to start daemons running at security contexts dependent on the remote client. Patch design approach: - Design for TCP The patch enables the SELinux LSM to set the peer security context for a socket based on the security context of the IPSec security association. The application may retrieve this context using getsockopt. When called, the kernel determines if the socket is a connected (TCP_ESTABLISHED) TCP socket and, if so, uses the dst_entry cache on the socket to retrieve the security associations. If a security association has a security context, the context string is returned, as for UNIX domain sockets. - Design for UDP Unlike TCP, UDP is connectionless. This requires a somewhat different API to retrieve the peer security context. With TCP, the peer security context stays the same throughout the connection, thus it can be retrieved at any time between when the connection is established and when it is torn down. With UDP, each read/write can have different peer and thus the security context might change every time. As a result the security context retrieval must be done TOGETHER with the packet retrieval. The solution is to build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). Patch implementation details: - Implementation for TCP The security context can be retrieved by applications using getsockopt with the existing SO_PEERSEC flag. As an example (ignoring error checking): getsockopt(sockfd, SOL_SOCKET, SO_PEERSEC, optbuf, &optlen); printf("Socket peer context is: %s\n", optbuf); The SELinux function, selinux_socket_getpeersec, is extended to check for labeled security associations for connected (TCP_ESTABLISHED == sk->sk_state) TCP sockets only. If so, the socket has a dst_cache of struct dst_entry values that may refer to security associations. If these have security associations with security contexts, the security context is returned. getsockopt returns a buffer that contains a security context string or the buffer is unmodified. - Implementation for UDP To retrieve the security context, the application first indicates to the kernel such desire by setting the IP_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for UDP should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_IP, IP_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_IP && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } ip_setsockopt is enhanced with a new socket option IP_PASSSEC to allow a server socket to receive security context of the peer. A new ancillary message type SCM_SECURITY. When the packet is received we get the security context from the sec_path pointer which is contained in the sk_buff, and copy it to the ancillary message space. An additional LSM hook, selinux_socket_getpeersec_udp, is defined to retrieve the security context from the SELinux space. The existing function, selinux_socket_getpeersec does not suit our purpose, because the security context is copied directly to user space, rather than to kernel space. Testing: We have tested the patch by setting up TCP and UDP connections between applications on two machines using the IPSec policies that result in labeled security associations being built. For TCP, we can then extract the peer security context using getsockopt on either end. For UDP, the receiving end can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: James Morris <jmorris@namei.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-21 14:41:23 +08:00
static int selinux_socket_getpeersec_stream(struct socket *sock, char __user *optval,
int __user *optlen, unsigned len)
{
int err = 0;
char *scontext;
u32 scontext_len;
struct sk_security_struct *ssec;
struct inode_security_struct *isec;
u32 peer_sid = SECSID_NULL;
isec = SOCK_INODE(sock)->i_security;
[SECURITY]: TCP/UDP getpeersec This patch implements an application of the LSM-IPSec networking controls whereby an application can determine the label of the security association its TCP or UDP sockets are currently connected to via getsockopt and the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of an IPSec security association a particular TCP or UDP socket is using. The application can then use this security context to determine the security context for processing on behalf of the peer at the other end of this connection. In the case of UDP, the security context is for each individual packet. An example application is the inetd daemon, which could be modified to start daemons running at security contexts dependent on the remote client. Patch design approach: - Design for TCP The patch enables the SELinux LSM to set the peer security context for a socket based on the security context of the IPSec security association. The application may retrieve this context using getsockopt. When called, the kernel determines if the socket is a connected (TCP_ESTABLISHED) TCP socket and, if so, uses the dst_entry cache on the socket to retrieve the security associations. If a security association has a security context, the context string is returned, as for UNIX domain sockets. - Design for UDP Unlike TCP, UDP is connectionless. This requires a somewhat different API to retrieve the peer security context. With TCP, the peer security context stays the same throughout the connection, thus it can be retrieved at any time between when the connection is established and when it is torn down. With UDP, each read/write can have different peer and thus the security context might change every time. As a result the security context retrieval must be done TOGETHER with the packet retrieval. The solution is to build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). Patch implementation details: - Implementation for TCP The security context can be retrieved by applications using getsockopt with the existing SO_PEERSEC flag. As an example (ignoring error checking): getsockopt(sockfd, SOL_SOCKET, SO_PEERSEC, optbuf, &optlen); printf("Socket peer context is: %s\n", optbuf); The SELinux function, selinux_socket_getpeersec, is extended to check for labeled security associations for connected (TCP_ESTABLISHED == sk->sk_state) TCP sockets only. If so, the socket has a dst_cache of struct dst_entry values that may refer to security associations. If these have security associations with security contexts, the security context is returned. getsockopt returns a buffer that contains a security context string or the buffer is unmodified. - Implementation for UDP To retrieve the security context, the application first indicates to the kernel such desire by setting the IP_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for UDP should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_IP, IP_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_IP && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } ip_setsockopt is enhanced with a new socket option IP_PASSSEC to allow a server socket to receive security context of the peer. A new ancillary message type SCM_SECURITY. When the packet is received we get the security context from the sec_path pointer which is contained in the sk_buff, and copy it to the ancillary message space. An additional LSM hook, selinux_socket_getpeersec_udp, is defined to retrieve the security context from the SELinux space. The existing function, selinux_socket_getpeersec does not suit our purpose, because the security context is copied directly to user space, rather than to kernel space. Testing: We have tested the patch by setting up TCP and UDP connections between applications on two machines using the IPSec policies that result in labeled security associations being built. For TCP, we can then extract the peer security context using getsockopt on either end. For UDP, the receiving end can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: James Morris <jmorris@namei.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-21 14:41:23 +08:00
if (isec->sclass == SECCLASS_UNIX_STREAM_SOCKET ||
isec->sclass == SECCLASS_TCP_SOCKET) {
[SECURITY]: TCP/UDP getpeersec This patch implements an application of the LSM-IPSec networking controls whereby an application can determine the label of the security association its TCP or UDP sockets are currently connected to via getsockopt and the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of an IPSec security association a particular TCP or UDP socket is using. The application can then use this security context to determine the security context for processing on behalf of the peer at the other end of this connection. In the case of UDP, the security context is for each individual packet. An example application is the inetd daemon, which could be modified to start daemons running at security contexts dependent on the remote client. Patch design approach: - Design for TCP The patch enables the SELinux LSM to set the peer security context for a socket based on the security context of the IPSec security association. The application may retrieve this context using getsockopt. When called, the kernel determines if the socket is a connected (TCP_ESTABLISHED) TCP socket and, if so, uses the dst_entry cache on the socket to retrieve the security associations. If a security association has a security context, the context string is returned, as for UNIX domain sockets. - Design for UDP Unlike TCP, UDP is connectionless. This requires a somewhat different API to retrieve the peer security context. With TCP, the peer security context stays the same throughout the connection, thus it can be retrieved at any time between when the connection is established and when it is torn down. With UDP, each read/write can have different peer and thus the security context might change every time. As a result the security context retrieval must be done TOGETHER with the packet retrieval. The solution is to build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). Patch implementation details: - Implementation for TCP The security context can be retrieved by applications using getsockopt with the existing SO_PEERSEC flag. As an example (ignoring error checking): getsockopt(sockfd, SOL_SOCKET, SO_PEERSEC, optbuf, &optlen); printf("Socket peer context is: %s\n", optbuf); The SELinux function, selinux_socket_getpeersec, is extended to check for labeled security associations for connected (TCP_ESTABLISHED == sk->sk_state) TCP sockets only. If so, the socket has a dst_cache of struct dst_entry values that may refer to security associations. If these have security associations with security contexts, the security context is returned. getsockopt returns a buffer that contains a security context string or the buffer is unmodified. - Implementation for UDP To retrieve the security context, the application first indicates to the kernel such desire by setting the IP_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for UDP should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_IP, IP_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_IP && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } ip_setsockopt is enhanced with a new socket option IP_PASSSEC to allow a server socket to receive security context of the peer. A new ancillary message type SCM_SECURITY. When the packet is received we get the security context from the sec_path pointer which is contained in the sk_buff, and copy it to the ancillary message space. An additional LSM hook, selinux_socket_getpeersec_udp, is defined to retrieve the security context from the SELinux space. The existing function, selinux_socket_getpeersec does not suit our purpose, because the security context is copied directly to user space, rather than to kernel space. Testing: We have tested the patch by setting up TCP and UDP connections between applications on two machines using the IPSec policies that result in labeled security associations being built. For TCP, we can then extract the peer security context using getsockopt on either end. For UDP, the receiving end can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: James Morris <jmorris@namei.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-21 14:41:23 +08:00
ssec = sock->sk->sk_security;
peer_sid = ssec->peer_sid;
}
if (peer_sid == SECSID_NULL) {
err = -ENOPROTOOPT;
goto out;
}
[SECURITY]: TCP/UDP getpeersec This patch implements an application of the LSM-IPSec networking controls whereby an application can determine the label of the security association its TCP or UDP sockets are currently connected to via getsockopt and the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of an IPSec security association a particular TCP or UDP socket is using. The application can then use this security context to determine the security context for processing on behalf of the peer at the other end of this connection. In the case of UDP, the security context is for each individual packet. An example application is the inetd daemon, which could be modified to start daemons running at security contexts dependent on the remote client. Patch design approach: - Design for TCP The patch enables the SELinux LSM to set the peer security context for a socket based on the security context of the IPSec security association. The application may retrieve this context using getsockopt. When called, the kernel determines if the socket is a connected (TCP_ESTABLISHED) TCP socket and, if so, uses the dst_entry cache on the socket to retrieve the security associations. If a security association has a security context, the context string is returned, as for UNIX domain sockets. - Design for UDP Unlike TCP, UDP is connectionless. This requires a somewhat different API to retrieve the peer security context. With TCP, the peer security context stays the same throughout the connection, thus it can be retrieved at any time between when the connection is established and when it is torn down. With UDP, each read/write can have different peer and thus the security context might change every time. As a result the security context retrieval must be done TOGETHER with the packet retrieval. The solution is to build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). Patch implementation details: - Implementation for TCP The security context can be retrieved by applications using getsockopt with the existing SO_PEERSEC flag. As an example (ignoring error checking): getsockopt(sockfd, SOL_SOCKET, SO_PEERSEC, optbuf, &optlen); printf("Socket peer context is: %s\n", optbuf); The SELinux function, selinux_socket_getpeersec, is extended to check for labeled security associations for connected (TCP_ESTABLISHED == sk->sk_state) TCP sockets only. If so, the socket has a dst_cache of struct dst_entry values that may refer to security associations. If these have security associations with security contexts, the security context is returned. getsockopt returns a buffer that contains a security context string or the buffer is unmodified. - Implementation for UDP To retrieve the security context, the application first indicates to the kernel such desire by setting the IP_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for UDP should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_IP, IP_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_IP && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } ip_setsockopt is enhanced with a new socket option IP_PASSSEC to allow a server socket to receive security context of the peer. A new ancillary message type SCM_SECURITY. When the packet is received we get the security context from the sec_path pointer which is contained in the sk_buff, and copy it to the ancillary message space. An additional LSM hook, selinux_socket_getpeersec_udp, is defined to retrieve the security context from the SELinux space. The existing function, selinux_socket_getpeersec does not suit our purpose, because the security context is copied directly to user space, rather than to kernel space. Testing: We have tested the patch by setting up TCP and UDP connections between applications on two machines using the IPSec policies that result in labeled security associations being built. For TCP, we can then extract the peer security context using getsockopt on either end. For UDP, the receiving end can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: James Morris <jmorris@namei.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-21 14:41:23 +08:00
err = security_sid_to_context(peer_sid, &scontext, &scontext_len);
if (err)
goto out;
if (scontext_len > len) {
err = -ERANGE;
goto out_len;
}
if (copy_to_user(optval, scontext, scontext_len))
err = -EFAULT;
out_len:
if (put_user(scontext_len, optlen))
err = -EFAULT;
kfree(scontext);
out:
return err;
}
static int selinux_socket_getpeersec_dgram(struct socket *sock, struct sk_buff *skb, u32 *secid)
[SECURITY]: TCP/UDP getpeersec This patch implements an application of the LSM-IPSec networking controls whereby an application can determine the label of the security association its TCP or UDP sockets are currently connected to via getsockopt and the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of an IPSec security association a particular TCP or UDP socket is using. The application can then use this security context to determine the security context for processing on behalf of the peer at the other end of this connection. In the case of UDP, the security context is for each individual packet. An example application is the inetd daemon, which could be modified to start daemons running at security contexts dependent on the remote client. Patch design approach: - Design for TCP The patch enables the SELinux LSM to set the peer security context for a socket based on the security context of the IPSec security association. The application may retrieve this context using getsockopt. When called, the kernel determines if the socket is a connected (TCP_ESTABLISHED) TCP socket and, if so, uses the dst_entry cache on the socket to retrieve the security associations. If a security association has a security context, the context string is returned, as for UNIX domain sockets. - Design for UDP Unlike TCP, UDP is connectionless. This requires a somewhat different API to retrieve the peer security context. With TCP, the peer security context stays the same throughout the connection, thus it can be retrieved at any time between when the connection is established and when it is torn down. With UDP, each read/write can have different peer and thus the security context might change every time. As a result the security context retrieval must be done TOGETHER with the packet retrieval. The solution is to build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). Patch implementation details: - Implementation for TCP The security context can be retrieved by applications using getsockopt with the existing SO_PEERSEC flag. As an example (ignoring error checking): getsockopt(sockfd, SOL_SOCKET, SO_PEERSEC, optbuf, &optlen); printf("Socket peer context is: %s\n", optbuf); The SELinux function, selinux_socket_getpeersec, is extended to check for labeled security associations for connected (TCP_ESTABLISHED == sk->sk_state) TCP sockets only. If so, the socket has a dst_cache of struct dst_entry values that may refer to security associations. If these have security associations with security contexts, the security context is returned. getsockopt returns a buffer that contains a security context string or the buffer is unmodified. - Implementation for UDP To retrieve the security context, the application first indicates to the kernel such desire by setting the IP_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for UDP should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_IP, IP_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_IP && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } ip_setsockopt is enhanced with a new socket option IP_PASSSEC to allow a server socket to receive security context of the peer. A new ancillary message type SCM_SECURITY. When the packet is received we get the security context from the sec_path pointer which is contained in the sk_buff, and copy it to the ancillary message space. An additional LSM hook, selinux_socket_getpeersec_udp, is defined to retrieve the security context from the SELinux space. The existing function, selinux_socket_getpeersec does not suit our purpose, because the security context is copied directly to user space, rather than to kernel space. Testing: We have tested the patch by setting up TCP and UDP connections between applications on two machines using the IPSec policies that result in labeled security associations being built. For TCP, we can then extract the peer security context using getsockopt on either end. For UDP, the receiving end can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: James Morris <jmorris@namei.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-21 14:41:23 +08:00
{
u32 peer_secid = SECSID_NULL;
u16 family;
[AF_UNIX]: Datagram getpeersec This patch implements an API whereby an application can determine the label of its peer's Unix datagram sockets via the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of the peer of a Unix datagram socket. The application can then use this security context to determine the security context for processing on behalf of the peer who sent the packet. Patch design and implementation: The design and implementation is very similar to the UDP case for INET sockets. Basically we build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). To retrieve the security context, the application first indicates to the kernel such desire by setting the SO_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for Unix datagram socket should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_SOCKET, SO_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_SOCKET && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } sock_setsockopt is enhanced with a new socket option SOCK_PASSSEC to allow a server socket to receive security context of the peer. Testing: We have tested the patch by setting up Unix datagram client and server applications. We verified that the server can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: Acked-by: James Morris <jmorris@namei.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-30 03:27:47 +08:00
if (skb && skb->protocol == htons(ETH_P_IP))
family = PF_INET;
else if (skb && skb->protocol == htons(ETH_P_IPV6))
family = PF_INET6;
else if (sock)
family = sock->sk->sk_family;
else
goto out;
if (sock && family == PF_UNIX)
selinux_inode_getsecid(SOCK_INODE(sock), &peer_secid);
else if (skb)
selinux_skb_peerlbl_sid(skb, family, &peer_secid);
[SECURITY]: TCP/UDP getpeersec This patch implements an application of the LSM-IPSec networking controls whereby an application can determine the label of the security association its TCP or UDP sockets are currently connected to via getsockopt and the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of an IPSec security association a particular TCP or UDP socket is using. The application can then use this security context to determine the security context for processing on behalf of the peer at the other end of this connection. In the case of UDP, the security context is for each individual packet. An example application is the inetd daemon, which could be modified to start daemons running at security contexts dependent on the remote client. Patch design approach: - Design for TCP The patch enables the SELinux LSM to set the peer security context for a socket based on the security context of the IPSec security association. The application may retrieve this context using getsockopt. When called, the kernel determines if the socket is a connected (TCP_ESTABLISHED) TCP socket and, if so, uses the dst_entry cache on the socket to retrieve the security associations. If a security association has a security context, the context string is returned, as for UNIX domain sockets. - Design for UDP Unlike TCP, UDP is connectionless. This requires a somewhat different API to retrieve the peer security context. With TCP, the peer security context stays the same throughout the connection, thus it can be retrieved at any time between when the connection is established and when it is torn down. With UDP, each read/write can have different peer and thus the security context might change every time. As a result the security context retrieval must be done TOGETHER with the packet retrieval. The solution is to build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). Patch implementation details: - Implementation for TCP The security context can be retrieved by applications using getsockopt with the existing SO_PEERSEC flag. As an example (ignoring error checking): getsockopt(sockfd, SOL_SOCKET, SO_PEERSEC, optbuf, &optlen); printf("Socket peer context is: %s\n", optbuf); The SELinux function, selinux_socket_getpeersec, is extended to check for labeled security associations for connected (TCP_ESTABLISHED == sk->sk_state) TCP sockets only. If so, the socket has a dst_cache of struct dst_entry values that may refer to security associations. If these have security associations with security contexts, the security context is returned. getsockopt returns a buffer that contains a security context string or the buffer is unmodified. - Implementation for UDP To retrieve the security context, the application first indicates to the kernel such desire by setting the IP_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for UDP should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_IP, IP_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_IP && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } ip_setsockopt is enhanced with a new socket option IP_PASSSEC to allow a server socket to receive security context of the peer. A new ancillary message type SCM_SECURITY. When the packet is received we get the security context from the sec_path pointer which is contained in the sk_buff, and copy it to the ancillary message space. An additional LSM hook, selinux_socket_getpeersec_udp, is defined to retrieve the security context from the SELinux space. The existing function, selinux_socket_getpeersec does not suit our purpose, because the security context is copied directly to user space, rather than to kernel space. Testing: We have tested the patch by setting up TCP and UDP connections between applications on two machines using the IPSec policies that result in labeled security associations being built. For TCP, we can then extract the peer security context using getsockopt on either end. For UDP, the receiving end can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: James Morris <jmorris@namei.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-21 14:41:23 +08:00
out:
*secid = peer_secid;
if (peer_secid == SECSID_NULL)
return -EINVAL;
return 0;
[SECURITY]: TCP/UDP getpeersec This patch implements an application of the LSM-IPSec networking controls whereby an application can determine the label of the security association its TCP or UDP sockets are currently connected to via getsockopt and the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of an IPSec security association a particular TCP or UDP socket is using. The application can then use this security context to determine the security context for processing on behalf of the peer at the other end of this connection. In the case of UDP, the security context is for each individual packet. An example application is the inetd daemon, which could be modified to start daemons running at security contexts dependent on the remote client. Patch design approach: - Design for TCP The patch enables the SELinux LSM to set the peer security context for a socket based on the security context of the IPSec security association. The application may retrieve this context using getsockopt. When called, the kernel determines if the socket is a connected (TCP_ESTABLISHED) TCP socket and, if so, uses the dst_entry cache on the socket to retrieve the security associations. If a security association has a security context, the context string is returned, as for UNIX domain sockets. - Design for UDP Unlike TCP, UDP is connectionless. This requires a somewhat different API to retrieve the peer security context. With TCP, the peer security context stays the same throughout the connection, thus it can be retrieved at any time between when the connection is established and when it is torn down. With UDP, each read/write can have different peer and thus the security context might change every time. As a result the security context retrieval must be done TOGETHER with the packet retrieval. The solution is to build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). Patch implementation details: - Implementation for TCP The security context can be retrieved by applications using getsockopt with the existing SO_PEERSEC flag. As an example (ignoring error checking): getsockopt(sockfd, SOL_SOCKET, SO_PEERSEC, optbuf, &optlen); printf("Socket peer context is: %s\n", optbuf); The SELinux function, selinux_socket_getpeersec, is extended to check for labeled security associations for connected (TCP_ESTABLISHED == sk->sk_state) TCP sockets only. If so, the socket has a dst_cache of struct dst_entry values that may refer to security associations. If these have security associations with security contexts, the security context is returned. getsockopt returns a buffer that contains a security context string or the buffer is unmodified. - Implementation for UDP To retrieve the security context, the application first indicates to the kernel such desire by setting the IP_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for UDP should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_IP, IP_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_IP && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } ip_setsockopt is enhanced with a new socket option IP_PASSSEC to allow a server socket to receive security context of the peer. A new ancillary message type SCM_SECURITY. When the packet is received we get the security context from the sec_path pointer which is contained in the sk_buff, and copy it to the ancillary message space. An additional LSM hook, selinux_socket_getpeersec_udp, is defined to retrieve the security context from the SELinux space. The existing function, selinux_socket_getpeersec does not suit our purpose, because the security context is copied directly to user space, rather than to kernel space. Testing: We have tested the patch by setting up TCP and UDP connections between applications on two machines using the IPSec policies that result in labeled security associations being built. For TCP, we can then extract the peer security context using getsockopt on either end. For UDP, the receiving end can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: James Morris <jmorris@namei.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-21 14:41:23 +08:00
}
static int selinux_sk_alloc_security(struct sock *sk, int family, gfp_t priority)
{
return sk_alloc_security(sk, family, priority);
}
static void selinux_sk_free_security(struct sock *sk)
{
sk_free_security(sk);
}
static void selinux_sk_clone_security(const struct sock *sk, struct sock *newsk)
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
{
struct sk_security_struct *ssec = sk->sk_security;
struct sk_security_struct *newssec = newsk->sk_security;
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
newssec->sid = ssec->sid;
newssec->peer_sid = ssec->peer_sid;
newssec->sclass = ssec->sclass;
selinux_netlbl_sk_security_reset(newssec, newsk->sk_family);
}
static void selinux_sk_getsecid(struct sock *sk, u32 *secid)
{
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
if (!sk)
*secid = SECINITSID_ANY_SOCKET;
else {
struct sk_security_struct *sksec = sk->sk_security;
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
*secid = sksec->sid;
}
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
}
static void selinux_sock_graft(struct sock *sk, struct socket *parent)
{
struct inode_security_struct *isec = SOCK_INODE(parent)->i_security;
struct sk_security_struct *sksec = sk->sk_security;
if (sk->sk_family == PF_INET || sk->sk_family == PF_INET6 ||
sk->sk_family == PF_UNIX)
isec->sid = sksec->sid;
sksec->sclass = isec->sclass;
}
static int selinux_inet_conn_request(struct sock *sk, struct sk_buff *skb,
struct request_sock *req)
{
struct sk_security_struct *sksec = sk->sk_security;
int err;
u16 family = sk->sk_family;
u32 newsid;
u32 peersid;
/* handle mapped IPv4 packets arriving via IPv6 sockets */
if (family == PF_INET6 && skb->protocol == htons(ETH_P_IP))
family = PF_INET;
err = selinux_skb_peerlbl_sid(skb, family, &peersid);
if (err)
return err;
if (peersid == SECSID_NULL) {
req->secid = sksec->sid;
req->peer_secid = SECSID_NULL;
return 0;
}
err = security_sid_mls_copy(sksec->sid, peersid, &newsid);
if (err)
return err;
req->secid = newsid;
req->peer_secid = peersid;
return 0;
}
static void selinux_inet_csk_clone(struct sock *newsk,
const struct request_sock *req)
{
struct sk_security_struct *newsksec = newsk->sk_security;
newsksec->sid = req->secid;
newsksec->peer_sid = req->peer_secid;
/* NOTE: Ideally, we should also get the isec->sid for the
new socket in sync, but we don't have the isec available yet.
So we will wait until sock_graft to do it, by which
time it will have been created and available. */
/* We don't need to take any sort of lock here as we are the only
* thread with access to newsksec */
selinux_netlbl_sk_security_reset(newsksec, req->rsk_ops->family);
}
static void selinux_inet_conn_established(struct sock *sk, struct sk_buff *skb)
{
u16 family = sk->sk_family;
struct sk_security_struct *sksec = sk->sk_security;
/* handle mapped IPv4 packets arriving via IPv6 sockets */
if (family == PF_INET6 && skb->protocol == htons(ETH_P_IP))
family = PF_INET;
selinux_skb_peerlbl_sid(skb, family, &sksec->peer_sid);
selinux_netlbl_inet_conn_established(sk, family);
}
static void selinux_req_classify_flow(const struct request_sock *req,
struct flowi *fl)
{
fl->secid = req->secid;
}
static int selinux_nlmsg_perm(struct sock *sk, struct sk_buff *skb)
{
int err = 0;
u32 perm;
struct nlmsghdr *nlh;
struct socket *sock = sk->sk_socket;
struct inode_security_struct *isec = SOCK_INODE(sock)->i_security;
if (skb->len < NLMSG_SPACE(0)) {
err = -EINVAL;
goto out;
}
nlh = nlmsg_hdr(skb);
err = selinux_nlmsg_lookup(isec->sclass, nlh->nlmsg_type, &perm);
if (err) {
if (err == -EINVAL) {
audit_log(current->audit_context, GFP_KERNEL, AUDIT_SELINUX_ERR,
"SELinux: unrecognized netlink message"
" type=%hu for sclass=%hu\n",
nlh->nlmsg_type, isec->sclass);
if (!selinux_enforcing)
err = 0;
}
/* Ignore */
if (err == -ENOENT)
err = 0;
goto out;
}
err = socket_has_perm(current, sock, perm);
out:
return err;
}
#ifdef CONFIG_NETFILTER
static unsigned int selinux_ip_forward(struct sk_buff *skb, int ifindex,
u16 family)
{
int err;
char *addrp;
u32 peer_sid;
struct avc_audit_data ad;
u8 secmark_active;
u8 netlbl_active;
u8 peerlbl_active;
if (!selinux_policycap_netpeer)
return NF_ACCEPT;
secmark_active = selinux_secmark_enabled();
netlbl_active = netlbl_enabled();
peerlbl_active = netlbl_active || selinux_xfrm_enabled();
if (!secmark_active && !peerlbl_active)
return NF_ACCEPT;
if (selinux_skb_peerlbl_sid(skb, family, &peer_sid) != 0)
return NF_DROP;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.netif = ifindex;
ad.u.net.family = family;
if (selinux_parse_skb(skb, &ad, &addrp, 1, NULL) != 0)
return NF_DROP;
if (peerlbl_active) {
err = selinux_inet_sys_rcv_skb(ifindex, addrp, family,
peer_sid, &ad);
if (err) {
selinux_netlbl_err(skb, err, 1);
return NF_DROP;
}
}
if (secmark_active)
if (avc_has_perm(peer_sid, skb->secmark,
SECCLASS_PACKET, PACKET__FORWARD_IN, &ad))
return NF_DROP;
if (netlbl_active)
/* we do this in the FORWARD path and not the POST_ROUTING
* path because we want to make sure we apply the necessary
* labeling before IPsec is applied so we can leverage AH
* protection */
if (selinux_netlbl_skbuff_setsid(skb, family, peer_sid) != 0)
return NF_DROP;
return NF_ACCEPT;
}
static unsigned int selinux_ipv4_forward(unsigned int hooknum,
struct sk_buff *skb,
const struct net_device *in,
const struct net_device *out,
int (*okfn)(struct sk_buff *))
{
return selinux_ip_forward(skb, in->ifindex, PF_INET);
}
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
static unsigned int selinux_ipv6_forward(unsigned int hooknum,
struct sk_buff *skb,
const struct net_device *in,
const struct net_device *out,
int (*okfn)(struct sk_buff *))
{
return selinux_ip_forward(skb, in->ifindex, PF_INET6);
}
#endif /* IPV6 */
static unsigned int selinux_ip_output(struct sk_buff *skb,
u16 family)
{
u32 sid;
if (!netlbl_enabled())
return NF_ACCEPT;
/* we do this in the LOCAL_OUT path and not the POST_ROUTING path
* because we want to make sure we apply the necessary labeling
* before IPsec is applied so we can leverage AH protection */
if (skb->sk) {
struct sk_security_struct *sksec = skb->sk->sk_security;
sid = sksec->sid;
} else
sid = SECINITSID_KERNEL;
if (selinux_netlbl_skbuff_setsid(skb, family, sid) != 0)
return NF_DROP;
return NF_ACCEPT;
}
static unsigned int selinux_ipv4_output(unsigned int hooknum,
struct sk_buff *skb,
const struct net_device *in,
const struct net_device *out,
int (*okfn)(struct sk_buff *))
{
return selinux_ip_output(skb, PF_INET);
}
static int selinux_ip_postroute_iptables_compat(struct sock *sk,
int ifindex,
struct avc_audit_data *ad,
u16 family, char *addrp)
{
int err;
struct sk_security_struct *sksec = sk->sk_security;
u16 sk_class;
u32 netif_perm, node_perm, send_perm;
u32 port_sid, node_sid, if_sid, sk_sid;
sk_sid = sksec->sid;
sk_class = sksec->sclass;
switch (sk_class) {
case SECCLASS_UDP_SOCKET:
netif_perm = NETIF__UDP_SEND;
node_perm = NODE__UDP_SEND;
send_perm = UDP_SOCKET__SEND_MSG;
break;
case SECCLASS_TCP_SOCKET:
netif_perm = NETIF__TCP_SEND;
node_perm = NODE__TCP_SEND;
send_perm = TCP_SOCKET__SEND_MSG;
break;
case SECCLASS_DCCP_SOCKET:
netif_perm = NETIF__DCCP_SEND;
node_perm = NODE__DCCP_SEND;
send_perm = DCCP_SOCKET__SEND_MSG;
break;
default:
netif_perm = NETIF__RAWIP_SEND;
node_perm = NODE__RAWIP_SEND;
send_perm = 0;
break;
}
err = sel_netif_sid(ifindex, &if_sid);
if (err)
return err;
err = avc_has_perm(sk_sid, if_sid, SECCLASS_NETIF, netif_perm, ad);
return err;
err = sel_netnode_sid(addrp, family, &node_sid);
if (err)
return err;
err = avc_has_perm(sk_sid, node_sid, SECCLASS_NODE, node_perm, ad);
if (err)
return err;
if (send_perm != 0)
return 0;
err = sel_netport_sid(sk->sk_protocol,
ntohs(ad->u.net.dport), &port_sid);
if (unlikely(err)) {
printk(KERN_WARNING
"SELinux: failure in"
" selinux_ip_postroute_iptables_compat(),"
" network port label not found\n");
return err;
}
return avc_has_perm(sk_sid, port_sid, sk_class, send_perm, ad);
}
static unsigned int selinux_ip_postroute_compat(struct sk_buff *skb,
int ifindex,
u16 family)
{
struct sock *sk = skb->sk;
struct sk_security_struct *sksec;
struct avc_audit_data ad;
char *addrp;
u8 proto;
if (sk == NULL)
return NF_ACCEPT;
sksec = sk->sk_security;
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.netif = ifindex;
ad.u.net.family = family;
if (selinux_parse_skb(skb, &ad, &addrp, 0, &proto))
return NF_DROP;
if (selinux_compat_net) {
if (selinux_ip_postroute_iptables_compat(skb->sk, ifindex,
&ad, family, addrp))
return NF_DROP;
} else {
if (avc_has_perm(sksec->sid, skb->secmark,
SECCLASS_PACKET, PACKET__SEND, &ad))
return NF_DROP;
}
if (selinux_policycap_netpeer)
if (selinux_xfrm_postroute_last(sksec->sid, skb, &ad, proto))
return NF_DROP;
return NF_ACCEPT;
}
static unsigned int selinux_ip_postroute(struct sk_buff *skb, int ifindex,
u16 family)
{
u32 secmark_perm;
u32 peer_sid;
struct sock *sk;
struct avc_audit_data ad;
char *addrp;
u8 secmark_active;
u8 peerlbl_active;
/* If any sort of compatibility mode is enabled then handoff processing
* to the selinux_ip_postroute_compat() function to deal with the
* special handling. We do this in an attempt to keep this function
* as fast and as clean as possible. */
if (selinux_compat_net || !selinux_policycap_netpeer)
return selinux_ip_postroute_compat(skb, ifindex, family);
/* If skb->dst->xfrm is non-NULL then the packet is undergoing an IPsec
* packet transformation so allow the packet to pass without any checks
* since we'll have another chance to perform access control checks
* when the packet is on it's final way out.
* NOTE: there appear to be some IPv6 multicast cases where skb->dst
* is NULL, in this case go ahead and apply access control. */
if (skb->dst != NULL && skb->dst->xfrm != NULL)
return NF_ACCEPT;
secmark_active = selinux_secmark_enabled();
peerlbl_active = netlbl_enabled() || selinux_xfrm_enabled();
if (!secmark_active && !peerlbl_active)
return NF_ACCEPT;
/* if the packet is being forwarded then get the peer label from the
* packet itself; otherwise check to see if it is from a local
* application or the kernel, if from an application get the peer label
* from the sending socket, otherwise use the kernel's sid */
sk = skb->sk;
if (sk == NULL) {
switch (family) {
case PF_INET:
if (IPCB(skb)->flags & IPSKB_FORWARDED)
secmark_perm = PACKET__FORWARD_OUT;
else
secmark_perm = PACKET__SEND;
break;
case PF_INET6:
if (IP6CB(skb)->flags & IP6SKB_FORWARDED)
secmark_perm = PACKET__FORWARD_OUT;
else
secmark_perm = PACKET__SEND;
break;
default:
return NF_DROP;
}
if (secmark_perm == PACKET__FORWARD_OUT) {
if (selinux_skb_peerlbl_sid(skb, family, &peer_sid))
return NF_DROP;
} else
peer_sid = SECINITSID_KERNEL;
} else {
struct sk_security_struct *sksec = sk->sk_security;
peer_sid = sksec->sid;
secmark_perm = PACKET__SEND;
}
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
AVC_AUDIT_DATA_INIT(&ad, NET);
ad.u.net.netif = ifindex;
ad.u.net.family = family;
if (selinux_parse_skb(skb, &ad, &addrp, 0, NULL))
return NF_DROP;
if (secmark_active)
if (avc_has_perm(peer_sid, skb->secmark,
SECCLASS_PACKET, secmark_perm, &ad))
return NF_DROP;
if (peerlbl_active) {
u32 if_sid;
u32 node_sid;
if (sel_netif_sid(ifindex, &if_sid))
return NF_DROP;
if (avc_has_perm(peer_sid, if_sid,
SECCLASS_NETIF, NETIF__EGRESS, &ad))
return NF_DROP;
if (sel_netnode_sid(addrp, family, &node_sid))
return NF_DROP;
if (avc_has_perm(peer_sid, node_sid,
SECCLASS_NODE, NODE__SENDTO, &ad))
return NF_DROP;
}
return NF_ACCEPT;
}
static unsigned int selinux_ipv4_postroute(unsigned int hooknum,
struct sk_buff *skb,
const struct net_device *in,
const struct net_device *out,
int (*okfn)(struct sk_buff *))
{
return selinux_ip_postroute(skb, out->ifindex, PF_INET);
}
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
static unsigned int selinux_ipv6_postroute(unsigned int hooknum,
struct sk_buff *skb,
const struct net_device *in,
const struct net_device *out,
int (*okfn)(struct sk_buff *))
{
return selinux_ip_postroute(skb, out->ifindex, PF_INET6);
}
#endif /* IPV6 */
#endif /* CONFIG_NETFILTER */
static int selinux_netlink_send(struct sock *sk, struct sk_buff *skb)
{
int err;
err = secondary_ops->netlink_send(sk, skb);
if (err)
return err;
if (policydb_loaded_version >= POLICYDB_VERSION_NLCLASS)
err = selinux_nlmsg_perm(sk, skb);
return err;
}
static int selinux_netlink_recv(struct sk_buff *skb, int capability)
{
int err;
struct avc_audit_data ad;
err = secondary_ops->netlink_recv(skb, capability);
if (err)
return err;
AVC_AUDIT_DATA_INIT(&ad, CAP);
ad.u.cap = capability;
return avc_has_perm(NETLINK_CB(skb).sid, NETLINK_CB(skb).sid,
SECCLASS_CAPABILITY, CAP_TO_MASK(capability), &ad);
}
static int ipc_alloc_security(struct task_struct *task,
struct kern_ipc_perm *perm,
u16 sclass)
{
struct task_security_struct *tsec = task->security;
struct ipc_security_struct *isec;
isec = kzalloc(sizeof(struct ipc_security_struct), GFP_KERNEL);
if (!isec)
return -ENOMEM;
isec->sclass = sclass;
isec->sid = tsec->sid;
perm->security = isec;
return 0;
}
static void ipc_free_security(struct kern_ipc_perm *perm)
{
struct ipc_security_struct *isec = perm->security;
perm->security = NULL;
kfree(isec);
}
static int msg_msg_alloc_security(struct msg_msg *msg)
{
struct msg_security_struct *msec;
msec = kzalloc(sizeof(struct msg_security_struct), GFP_KERNEL);
if (!msec)
return -ENOMEM;
msec->sid = SECINITSID_UNLABELED;
msg->security = msec;
return 0;
}
static void msg_msg_free_security(struct msg_msg *msg)
{
struct msg_security_struct *msec = msg->security;
msg->security = NULL;
kfree(msec);
}
static int ipc_has_perm(struct kern_ipc_perm *ipc_perms,
u32 perms)
{
struct task_security_struct *tsec;
struct ipc_security_struct *isec;
struct avc_audit_data ad;
tsec = current->security;
isec = ipc_perms->security;
AVC_AUDIT_DATA_INIT(&ad, IPC);
ad.u.ipc_id = ipc_perms->key;
return avc_has_perm(tsec->sid, isec->sid, isec->sclass, perms, &ad);
}
static int selinux_msg_msg_alloc_security(struct msg_msg *msg)
{
return msg_msg_alloc_security(msg);
}
static void selinux_msg_msg_free_security(struct msg_msg *msg)
{
msg_msg_free_security(msg);
}
/* message queue security operations */
static int selinux_msg_queue_alloc_security(struct msg_queue *msq)
{
struct task_security_struct *tsec;
struct ipc_security_struct *isec;
struct avc_audit_data ad;
int rc;
rc = ipc_alloc_security(current, &msq->q_perm, SECCLASS_MSGQ);
if (rc)
return rc;
tsec = current->security;
isec = msq->q_perm.security;
AVC_AUDIT_DATA_INIT(&ad, IPC);
ad.u.ipc_id = msq->q_perm.key;
rc = avc_has_perm(tsec->sid, isec->sid, SECCLASS_MSGQ,
MSGQ__CREATE, &ad);
if (rc) {
ipc_free_security(&msq->q_perm);
return rc;
}
return 0;
}
static void selinux_msg_queue_free_security(struct msg_queue *msq)
{
ipc_free_security(&msq->q_perm);
}
static int selinux_msg_queue_associate(struct msg_queue *msq, int msqflg)
{
struct task_security_struct *tsec;
struct ipc_security_struct *isec;
struct avc_audit_data ad;
tsec = current->security;
isec = msq->q_perm.security;
AVC_AUDIT_DATA_INIT(&ad, IPC);
ad.u.ipc_id = msq->q_perm.key;
return avc_has_perm(tsec->sid, isec->sid, SECCLASS_MSGQ,
MSGQ__ASSOCIATE, &ad);
}
static int selinux_msg_queue_msgctl(struct msg_queue *msq, int cmd)
{
int err;
int perms;
switch (cmd) {
case IPC_INFO:
case MSG_INFO:
/* No specific object, just general system-wide information. */
return task_has_system(current, SYSTEM__IPC_INFO);
case IPC_STAT:
case MSG_STAT:
perms = MSGQ__GETATTR | MSGQ__ASSOCIATE;
break;
case IPC_SET:
perms = MSGQ__SETATTR;
break;
case IPC_RMID:
perms = MSGQ__DESTROY;
break;
default:
return 0;
}
err = ipc_has_perm(&msq->q_perm, perms);
return err;
}
static int selinux_msg_queue_msgsnd(struct msg_queue *msq, struct msg_msg *msg, int msqflg)
{
struct task_security_struct *tsec;
struct ipc_security_struct *isec;
struct msg_security_struct *msec;
struct avc_audit_data ad;
int rc;
tsec = current->security;
isec = msq->q_perm.security;
msec = msg->security;
/*
* First time through, need to assign label to the message
*/
if (msec->sid == SECINITSID_UNLABELED) {
/*
* Compute new sid based on current process and
* message queue this message will be stored in
*/
rc = security_transition_sid(tsec->sid,
isec->sid,
SECCLASS_MSG,
&msec->sid);
if (rc)
return rc;
}
AVC_AUDIT_DATA_INIT(&ad, IPC);
ad.u.ipc_id = msq->q_perm.key;
/* Can this process write to the queue? */
rc = avc_has_perm(tsec->sid, isec->sid, SECCLASS_MSGQ,
MSGQ__WRITE, &ad);
if (!rc)
/* Can this process send the message */
rc = avc_has_perm(tsec->sid, msec->sid,
SECCLASS_MSG, MSG__SEND, &ad);
if (!rc)
/* Can the message be put in the queue? */
rc = avc_has_perm(msec->sid, isec->sid,
SECCLASS_MSGQ, MSGQ__ENQUEUE, &ad);
return rc;
}
static int selinux_msg_queue_msgrcv(struct msg_queue *msq, struct msg_msg *msg,
struct task_struct *target,
long type, int mode)
{
struct task_security_struct *tsec;
struct ipc_security_struct *isec;
struct msg_security_struct *msec;
struct avc_audit_data ad;
int rc;
tsec = target->security;
isec = msq->q_perm.security;
msec = msg->security;
AVC_AUDIT_DATA_INIT(&ad, IPC);
ad.u.ipc_id = msq->q_perm.key;
rc = avc_has_perm(tsec->sid, isec->sid,
SECCLASS_MSGQ, MSGQ__READ, &ad);
if (!rc)
rc = avc_has_perm(tsec->sid, msec->sid,
SECCLASS_MSG, MSG__RECEIVE, &ad);
return rc;
}
/* Shared Memory security operations */
static int selinux_shm_alloc_security(struct shmid_kernel *shp)
{
struct task_security_struct *tsec;
struct ipc_security_struct *isec;
struct avc_audit_data ad;
int rc;
rc = ipc_alloc_security(current, &shp->shm_perm, SECCLASS_SHM);
if (rc)
return rc;
tsec = current->security;
isec = shp->shm_perm.security;
AVC_AUDIT_DATA_INIT(&ad, IPC);
ad.u.ipc_id = shp->shm_perm.key;
rc = avc_has_perm(tsec->sid, isec->sid, SECCLASS_SHM,
SHM__CREATE, &ad);
if (rc) {
ipc_free_security(&shp->shm_perm);
return rc;
}
return 0;
}
static void selinux_shm_free_security(struct shmid_kernel *shp)
{
ipc_free_security(&shp->shm_perm);
}
static int selinux_shm_associate(struct shmid_kernel *shp, int shmflg)
{
struct task_security_struct *tsec;
struct ipc_security_struct *isec;
struct avc_audit_data ad;
tsec = current->security;
isec = shp->shm_perm.security;
AVC_AUDIT_DATA_INIT(&ad, IPC);
ad.u.ipc_id = shp->shm_perm.key;
return avc_has_perm(tsec->sid, isec->sid, SECCLASS_SHM,
SHM__ASSOCIATE, &ad);
}
/* Note, at this point, shp is locked down */
static int selinux_shm_shmctl(struct shmid_kernel *shp, int cmd)
{
int perms;
int err;
switch (cmd) {
case IPC_INFO:
case SHM_INFO:
/* No specific object, just general system-wide information. */
return task_has_system(current, SYSTEM__IPC_INFO);
case IPC_STAT:
case SHM_STAT:
perms = SHM__GETATTR | SHM__ASSOCIATE;
break;
case IPC_SET:
perms = SHM__SETATTR;
break;
case SHM_LOCK:
case SHM_UNLOCK:
perms = SHM__LOCK;
break;
case IPC_RMID:
perms = SHM__DESTROY;
break;
default:
return 0;
}
err = ipc_has_perm(&shp->shm_perm, perms);
return err;
}
static int selinux_shm_shmat(struct shmid_kernel *shp,
char __user *shmaddr, int shmflg)
{
u32 perms;
int rc;
rc = secondary_ops->shm_shmat(shp, shmaddr, shmflg);
if (rc)
return rc;
if (shmflg & SHM_RDONLY)
perms = SHM__READ;
else
perms = SHM__READ | SHM__WRITE;
return ipc_has_perm(&shp->shm_perm, perms);
}
/* Semaphore security operations */
static int selinux_sem_alloc_security(struct sem_array *sma)
{
struct task_security_struct *tsec;
struct ipc_security_struct *isec;
struct avc_audit_data ad;
int rc;
rc = ipc_alloc_security(current, &sma->sem_perm, SECCLASS_SEM);
if (rc)
return rc;
tsec = current->security;
isec = sma->sem_perm.security;
AVC_AUDIT_DATA_INIT(&ad, IPC);
ad.u.ipc_id = sma->sem_perm.key;
rc = avc_has_perm(tsec->sid, isec->sid, SECCLASS_SEM,
SEM__CREATE, &ad);
if (rc) {
ipc_free_security(&sma->sem_perm);
return rc;
}
return 0;
}
static void selinux_sem_free_security(struct sem_array *sma)
{
ipc_free_security(&sma->sem_perm);
}
static int selinux_sem_associate(struct sem_array *sma, int semflg)
{
struct task_security_struct *tsec;
struct ipc_security_struct *isec;
struct avc_audit_data ad;
tsec = current->security;
isec = sma->sem_perm.security;
AVC_AUDIT_DATA_INIT(&ad, IPC);
ad.u.ipc_id = sma->sem_perm.key;
return avc_has_perm(tsec->sid, isec->sid, SECCLASS_SEM,
SEM__ASSOCIATE, &ad);
}
/* Note, at this point, sma is locked down */
static int selinux_sem_semctl(struct sem_array *sma, int cmd)
{
int err;
u32 perms;
switch (cmd) {
case IPC_INFO:
case SEM_INFO:
/* No specific object, just general system-wide information. */
return task_has_system(current, SYSTEM__IPC_INFO);
case GETPID:
case GETNCNT:
case GETZCNT:
perms = SEM__GETATTR;
break;
case GETVAL:
case GETALL:
perms = SEM__READ;
break;
case SETVAL:
case SETALL:
perms = SEM__WRITE;
break;
case IPC_RMID:
perms = SEM__DESTROY;
break;
case IPC_SET:
perms = SEM__SETATTR;
break;
case IPC_STAT:
case SEM_STAT:
perms = SEM__GETATTR | SEM__ASSOCIATE;
break;
default:
return 0;
}
err = ipc_has_perm(&sma->sem_perm, perms);
return err;
}
static int selinux_sem_semop(struct sem_array *sma,
struct sembuf *sops, unsigned nsops, int alter)
{
u32 perms;
if (alter)
perms = SEM__READ | SEM__WRITE;
else
perms = SEM__READ;
return ipc_has_perm(&sma->sem_perm, perms);
}
static int selinux_ipc_permission(struct kern_ipc_perm *ipcp, short flag)
{
u32 av = 0;
av = 0;
if (flag & S_IRUGO)
av |= IPC__UNIX_READ;
if (flag & S_IWUGO)
av |= IPC__UNIX_WRITE;
if (av == 0)
return 0;
return ipc_has_perm(ipcp, av);
}
static void selinux_ipc_getsecid(struct kern_ipc_perm *ipcp, u32 *secid)
{
struct ipc_security_struct *isec = ipcp->security;
*secid = isec->sid;
}
static void selinux_d_instantiate(struct dentry *dentry, struct inode *inode)
{
if (inode)
inode_doinit_with_dentry(inode, dentry);
}
static int selinux_getprocattr(struct task_struct *p,
char *name, char **value)
{
struct task_security_struct *tsec;
u32 sid;
int error;
unsigned len;
if (current != p) {
error = task_has_perm(current, p, PROCESS__GETATTR);
if (error)
return error;
}
tsec = p->security;
if (!strcmp(name, "current"))
sid = tsec->sid;
else if (!strcmp(name, "prev"))
sid = tsec->osid;
else if (!strcmp(name, "exec"))
sid = tsec->exec_sid;
else if (!strcmp(name, "fscreate"))
sid = tsec->create_sid;
else if (!strcmp(name, "keycreate"))
sid = tsec->keycreate_sid;
else if (!strcmp(name, "sockcreate"))
sid = tsec->sockcreate_sid;
else
return -EINVAL;
if (!sid)
return 0;
error = security_sid_to_context(sid, value, &len);
if (error)
return error;
return len;
}
static int selinux_setprocattr(struct task_struct *p,
char *name, void *value, size_t size)
{
struct task_security_struct *tsec;
struct task_struct *tracer;
u32 sid = 0;
int error;
char *str = value;
if (current != p) {
/* SELinux only allows a process to change its own
security attributes. */
return -EACCES;
}
/*
* Basic control over ability to set these attributes at all.
* current == p, but we'll pass them separately in case the
* above restriction is ever removed.
*/
if (!strcmp(name, "exec"))
error = task_has_perm(current, p, PROCESS__SETEXEC);
else if (!strcmp(name, "fscreate"))
error = task_has_perm(current, p, PROCESS__SETFSCREATE);
else if (!strcmp(name, "keycreate"))
error = task_has_perm(current, p, PROCESS__SETKEYCREATE);
else if (!strcmp(name, "sockcreate"))
error = task_has_perm(current, p, PROCESS__SETSOCKCREATE);
else if (!strcmp(name, "current"))
error = task_has_perm(current, p, PROCESS__SETCURRENT);
else
error = -EINVAL;
if (error)
return error;
/* Obtain a SID for the context, if one was specified. */
if (size && str[1] && str[1] != '\n') {
if (str[size-1] == '\n') {
str[size-1] = 0;
size--;
}
error = security_context_to_sid(value, size, &sid);
selinux: support deferred mapping of contexts Introduce SELinux support for deferred mapping of security contexts in the SID table upon policy reload, and use this support for inode security contexts when the context is not yet valid under the current policy. Only processes with CAP_MAC_ADMIN + mac_admin permission in policy can set undefined security contexts on inodes. Inodes with such undefined contexts are treated as having the unlabeled context until the context becomes valid upon a policy reload that defines the context. Context invalidation upon policy reload also uses this support to save the context information in the SID table and later recover it upon a subsequent policy reload that defines the context again. This support is to enable package managers and similar programs to set down file contexts unknown to the system policy at the time the file is created in order to better support placing loadable policy modules in packages and to support build systems that need to create images of different distro releases with different policies w/o requiring all of the contexts to be defined or legal in the build host policy. With this patch applied, the following sequence is possible, although in practice it is recommended that this permission only be allowed to specific program domains such as the package manager. # rmdir baz # rm bar # touch bar # chcon -t foo_exec_t bar # foo_exec_t is not yet defined chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument # cat setundefined.te policy_module(setundefined, 1.0) require { type unconfined_t; type unlabeled_t; } files_type(unlabeled_t) allow unconfined_t self:capability2 mac_admin; # make -f /usr/share/selinux/devel/Makefile setundefined.pp # semodule -i setundefined.pp # chcon -t foo_exec_t bar # foo_exec_t is not yet defined # mkdir -Z system_u:object_r:foo_exec_t baz # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # cat foo.te policy_module(foo, 1.0) type foo_exec_t; files_type(foo_exec_t) # make -f /usr/share/selinux/devel/Makefile foo.pp # semodule -i foo.pp # defines foo_exec_t # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r foo # ls -Zd bar baz -rw-r--r-- root root system_u:object_r:unlabeled_t bar drwxr-xr-x root root system_u:object_r:unlabeled_t baz # semodule -i foo.pp # ls -Zd bar baz -rw-r--r-- root root user_u:object_r:foo_exec_t bar drwxr-xr-x root root system_u:object_r:foo_exec_t baz # semodule -r setundefined foo # chcon -t foo_exec_t bar # no longer defined and not allowed chcon: failed to change context of `bar' to `system_u:object_r:foo_exec_t': Invalid argument # rmdir baz # mkdir -Z system_u:object_r:foo_exec_t baz mkdir: failed to set default file creation context to `system_u:object_r:foo_exec_t': Invalid argument Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2008-05-08 01:03:20 +08:00
if (error == -EINVAL && !strcmp(name, "fscreate")) {
if (!capable(CAP_MAC_ADMIN))
return error;
error = security_context_to_sid_force(value, size,
&sid);
}
if (error)
return error;
}
/* Permission checking based on the specified context is
performed during the actual operation (execve,
open/mkdir/...), when we know the full context of the
operation. See selinux_bprm_set_security for the execve
checks and may_create for the file creation checks. The
operation will then fail if the context is not permitted. */
tsec = p->security;
if (!strcmp(name, "exec"))
tsec->exec_sid = sid;
else if (!strcmp(name, "fscreate"))
tsec->create_sid = sid;
else if (!strcmp(name, "keycreate")) {
error = may_create_key(sid, p);
if (error)
return error;
tsec->keycreate_sid = sid;
} else if (!strcmp(name, "sockcreate"))
tsec->sockcreate_sid = sid;
else if (!strcmp(name, "current")) {
struct av_decision avd;
if (sid == 0)
return -EINVAL;
SELinux: add boundary support and thread context assignment The purpose of this patch is to assign per-thread security context under a constraint. It enables multi-threaded server application to kick a request handler with its fair security context, and helps some of userspace object managers to handle user's request. When we assign a per-thread security context, it must not have wider permissions than the original one. Because a multi-threaded process shares a single local memory, an arbitary per-thread security context also means another thread can easily refer violated information. The constraint on a per-thread security context requires a new domain has to be equal or weaker than its original one, when it tries to assign a per-thread security context. Bounds relationship between two types is a way to ensure a domain can never have wider permission than its bounds. We can define it in two explicit or implicit ways. The first way is using new TYPEBOUNDS statement. It enables to define a boundary of types explicitly. The other one expand the concept of existing named based hierarchy. If we defines a type with "." separated name like "httpd_t.php", toolchain implicitly set its bounds on "httpd_t". This feature requires a new policy version. The 24th version (POLICYDB_VERSION_BOUNDARY) enables to ship them into kernel space, and the following patch enables to handle it. Signed-off-by: KaiGai Kohei <kaigai@ak.jp.nec.com> Acked-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2008-08-28 15:35:57 +08:00
/*
* SELinux allows to change context in the following case only.
* - Single threaded processes.
* - Multi threaded processes intend to change its context into
* more restricted domain (defined by TYPEBOUNDS statement).
*/
if (atomic_read(&p->mm->mm_users) != 1) {
struct task_struct *g, *t;
struct mm_struct *mm = p->mm;
read_lock(&tasklist_lock);
do_each_thread(g, t) {
if (t->mm == mm && t != p) {
read_unlock(&tasklist_lock);
SELinux: add boundary support and thread context assignment The purpose of this patch is to assign per-thread security context under a constraint. It enables multi-threaded server application to kick a request handler with its fair security context, and helps some of userspace object managers to handle user's request. When we assign a per-thread security context, it must not have wider permissions than the original one. Because a multi-threaded process shares a single local memory, an arbitary per-thread security context also means another thread can easily refer violated information. The constraint on a per-thread security context requires a new domain has to be equal or weaker than its original one, when it tries to assign a per-thread security context. Bounds relationship between two types is a way to ensure a domain can never have wider permission than its bounds. We can define it in two explicit or implicit ways. The first way is using new TYPEBOUNDS statement. It enables to define a boundary of types explicitly. The other one expand the concept of existing named based hierarchy. If we defines a type with "." separated name like "httpd_t.php", toolchain implicitly set its bounds on "httpd_t". This feature requires a new policy version. The 24th version (POLICYDB_VERSION_BOUNDARY) enables to ship them into kernel space, and the following patch enables to handle it. Signed-off-by: KaiGai Kohei <kaigai@ak.jp.nec.com> Acked-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2008-08-28 15:35:57 +08:00
error = security_bounded_transition(tsec->sid, sid);
if (!error)
goto boundary_ok;
return error;
}
} while_each_thread(g, t);
read_unlock(&tasklist_lock);
}
SELinux: add boundary support and thread context assignment The purpose of this patch is to assign per-thread security context under a constraint. It enables multi-threaded server application to kick a request handler with its fair security context, and helps some of userspace object managers to handle user's request. When we assign a per-thread security context, it must not have wider permissions than the original one. Because a multi-threaded process shares a single local memory, an arbitary per-thread security context also means another thread can easily refer violated information. The constraint on a per-thread security context requires a new domain has to be equal or weaker than its original one, when it tries to assign a per-thread security context. Bounds relationship between two types is a way to ensure a domain can never have wider permission than its bounds. We can define it in two explicit or implicit ways. The first way is using new TYPEBOUNDS statement. It enables to define a boundary of types explicitly. The other one expand the concept of existing named based hierarchy. If we defines a type with "." separated name like "httpd_t.php", toolchain implicitly set its bounds on "httpd_t". This feature requires a new policy version. The 24th version (POLICYDB_VERSION_BOUNDARY) enables to ship them into kernel space, and the following patch enables to handle it. Signed-off-by: KaiGai Kohei <kaigai@ak.jp.nec.com> Acked-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2008-08-28 15:35:57 +08:00
boundary_ok:
/* Check permissions for the transition. */
error = avc_has_perm(tsec->sid, sid, SECCLASS_PROCESS,
PROCESS__DYNTRANSITION, NULL);
if (error)
return error;
/* Check for ptracing, and update the task SID if ok.
Otherwise, leave SID unchanged and fail. */
task_lock(p);
rcu_read_lock();
tracer = tracehook_tracer_task(p);
if (tracer != NULL) {
struct task_security_struct *ptsec = tracer->security;
u32 ptsid = ptsec->sid;
rcu_read_unlock();
error = avc_has_perm_noaudit(ptsid, sid,
SECCLASS_PROCESS,
PROCESS__PTRACE, 0, &avd);
if (!error)
tsec->sid = sid;
task_unlock(p);
avc_audit(ptsid, sid, SECCLASS_PROCESS,
PROCESS__PTRACE, &avd, error, NULL);
if (error)
return error;
} else {
rcu_read_unlock();
tsec->sid = sid;
task_unlock(p);
}
} else
return -EINVAL;
return size;
}
static int selinux_secid_to_secctx(u32 secid, char **secdata, u32 *seclen)
{
return security_sid_to_context(secid, secdata, seclen);
}
static int selinux_secctx_to_secid(const char *secdata, u32 seclen, u32 *secid)
{
return security_context_to_sid(secdata, seclen, secid);
}
static void selinux_release_secctx(char *secdata, u32 seclen)
{
kfree(secdata);
}
#ifdef CONFIG_KEYS
static int selinux_key_alloc(struct key *k, struct task_struct *tsk,
unsigned long flags)
{
struct task_security_struct *tsec = tsk->security;
struct key_security_struct *ksec;
ksec = kzalloc(sizeof(struct key_security_struct), GFP_KERNEL);
if (!ksec)
return -ENOMEM;
if (tsec->keycreate_sid)
ksec->sid = tsec->keycreate_sid;
else
ksec->sid = tsec->sid;
k->security = ksec;
return 0;
}
static void selinux_key_free(struct key *k)
{
struct key_security_struct *ksec = k->security;
k->security = NULL;
kfree(ksec);
}
static int selinux_key_permission(key_ref_t key_ref,
struct task_struct *ctx,
key_perm_t perm)
{
struct key *key;
struct task_security_struct *tsec;
struct key_security_struct *ksec;
key = key_ref_to_ptr(key_ref);
tsec = ctx->security;
ksec = key->security;
/* if no specific permissions are requested, we skip the
permission check. No serious, additional covert channels
appear to be created. */
if (perm == 0)
return 0;
return avc_has_perm(tsec->sid, ksec->sid,
SECCLASS_KEY, perm, NULL);
}
static int selinux_key_getsecurity(struct key *key, char **_buffer)
{
struct key_security_struct *ksec = key->security;
char *context = NULL;
unsigned len;
int rc;
rc = security_sid_to_context(ksec->sid, &context, &len);
if (!rc)
rc = len;
*_buffer = context;
return rc;
}
#endif
static struct security_operations selinux_ops = {
.name = "selinux",
security: Fix setting of PF_SUPERPRIV by __capable() Fix the setting of PF_SUPERPRIV by __capable() as it could corrupt the flags the target process if that is not the current process and it is trying to change its own flags in a different way at the same time. __capable() is using neither atomic ops nor locking to protect t->flags. This patch removes __capable() and introduces has_capability() that doesn't set PF_SUPERPRIV on the process being queried. This patch further splits security_ptrace() in two: (1) security_ptrace_may_access(). This passes judgement on whether one process may access another only (PTRACE_MODE_ATTACH for ptrace() and PTRACE_MODE_READ for /proc), and takes a pointer to the child process. current is the parent. (2) security_ptrace_traceme(). This passes judgement on PTRACE_TRACEME only, and takes only a pointer to the parent process. current is the child. In Smack and commoncap, this uses has_capability() to determine whether the parent will be permitted to use PTRACE_ATTACH if normal checks fail. This does not set PF_SUPERPRIV. Two of the instances of __capable() actually only act on current, and so have been changed to calls to capable(). Of the places that were using __capable(): (1) The OOM killer calls __capable() thrice when weighing the killability of a process. All of these now use has_capability(). (2) cap_ptrace() and smack_ptrace() were using __capable() to check to see whether the parent was allowed to trace any process. As mentioned above, these have been split. For PTRACE_ATTACH and /proc, capable() is now used, and for PTRACE_TRACEME, has_capability() is used. (3) cap_safe_nice() only ever saw current, so now uses capable(). (4) smack_setprocattr() rejected accesses to tasks other than current just after calling __capable(), so the order of these two tests have been switched and capable() is used instead. (5) In smack_file_send_sigiotask(), we need to allow privileged processes to receive SIGIO on files they're manipulating. (6) In smack_task_wait(), we let a process wait for a privileged process, whether or not the process doing the waiting is privileged. I've tested this with the LTP SELinux and syscalls testscripts. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Serge Hallyn <serue@us.ibm.com> Acked-by: Casey Schaufler <casey@schaufler-ca.com> Acked-by: Andrew G. Morgan <morgan@kernel.org> Acked-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: James Morris <jmorris@namei.org>
2008-08-14 18:37:28 +08:00
.ptrace_may_access = selinux_ptrace_may_access,
.ptrace_traceme = selinux_ptrace_traceme,
.capget = selinux_capget,
.capset_check = selinux_capset_check,
.capset_set = selinux_capset_set,
.sysctl = selinux_sysctl,
.capable = selinux_capable,
.quotactl = selinux_quotactl,
.quota_on = selinux_quota_on,
.syslog = selinux_syslog,
.vm_enough_memory = selinux_vm_enough_memory,
.netlink_send = selinux_netlink_send,
.netlink_recv = selinux_netlink_recv,
.bprm_alloc_security = selinux_bprm_alloc_security,
.bprm_free_security = selinux_bprm_free_security,
.bprm_apply_creds = selinux_bprm_apply_creds,
.bprm_post_apply_creds = selinux_bprm_post_apply_creds,
.bprm_set_security = selinux_bprm_set_security,
.bprm_check_security = selinux_bprm_check_security,
.bprm_secureexec = selinux_bprm_secureexec,
.sb_alloc_security = selinux_sb_alloc_security,
.sb_free_security = selinux_sb_free_security,
.sb_copy_data = selinux_sb_copy_data,
.sb_kern_mount = selinux_sb_kern_mount,
.sb_show_options = selinux_sb_show_options,
.sb_statfs = selinux_sb_statfs,
.sb_mount = selinux_mount,
.sb_umount = selinux_umount,
Security: add get, set, and cloning of superblock security information Adds security_get_sb_mnt_opts, security_set_sb_mnt_opts, and security_clont_sb_mnt_opts to the LSM and to SELinux. This will allow filesystems to directly own and control all of their mount options if they so choose. This interface deals only with option identifiers and strings so it should generic enough for any LSM which may come in the future. Filesystems which pass text mount data around in the kernel (almost all of them) need not currently make use of this interface when dealing with SELinux since it will still parse those strings as it always has. I assume future LSM's would do the same. NFS is the primary FS which does not use text mount data and thus must make use of this interface. An LSM would need to implement these functions only if they had mount time options, such as selinux has context= or fscontext=. If the LSM has no mount time options they could simply not implement and let the dummy ops take care of things. An LSM other than SELinux would need to define new option numbers in security.h and any FS which decides to own there own security options would need to be patched to use this new interface for every possible LSM. This is because it was stated to me very clearly that LSM's should not attempt to understand FS mount data and the burdon to understand security should be in the FS which owns the options. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Stephen D. Smalley <sds@tycho.nsa.gov> Signed-off-by: James Morris <jmorris@namei.org>
2007-12-01 02:00:35 +08:00
.sb_set_mnt_opts = selinux_set_mnt_opts,
.sb_clone_mnt_opts = selinux_sb_clone_mnt_opts,
.sb_parse_opts_str = selinux_parse_opts_str,
.inode_alloc_security = selinux_inode_alloc_security,
.inode_free_security = selinux_inode_free_security,
[PATCH] security: enable atomic inode security labeling The following patch set enables atomic security labeling of newly created inodes by altering the fs code to invoke a new LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state during the inode creation transaction. This parallels the existing processing for setting ACLs on newly created inodes. Otherwise, it is possible for new inodes to be accessed by another thread via the dcache prior to complete security setup (presently handled by the post_create/mkdir/... LSM hooks in the VFS) and a newly created inode may be left unlabeled on the disk in the event of a crash. SELinux presently works around the issue by ensuring that the incore inode security label is initialized to a special SID that is inaccessible to unprivileged processes (in accordance with policy), thereby preventing inappropriate access but potentially causing false denials on legitimate accesses. A simple test program demonstrates such false denials on SELinux, and the patch solves the problem. Similar such false denials have been encountered in real applications. This patch defines a new inode_init_security LSM hook to obtain the security attribute to apply to a newly created inode and to set up the incore inode security state for it, and adds a corresponding hook function implementation to SELinux. Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-09-10 04:01:35 +08:00
.inode_init_security = selinux_inode_init_security,
.inode_create = selinux_inode_create,
.inode_link = selinux_inode_link,
.inode_unlink = selinux_inode_unlink,
.inode_symlink = selinux_inode_symlink,
.inode_mkdir = selinux_inode_mkdir,
.inode_rmdir = selinux_inode_rmdir,
.inode_mknod = selinux_inode_mknod,
.inode_rename = selinux_inode_rename,
.inode_readlink = selinux_inode_readlink,
.inode_follow_link = selinux_inode_follow_link,
.inode_permission = selinux_inode_permission,
.inode_setattr = selinux_inode_setattr,
.inode_getattr = selinux_inode_getattr,
.inode_setxattr = selinux_inode_setxattr,
.inode_post_setxattr = selinux_inode_post_setxattr,
.inode_getxattr = selinux_inode_getxattr,
.inode_listxattr = selinux_inode_listxattr,
.inode_removexattr = selinux_inode_removexattr,
.inode_getsecurity = selinux_inode_getsecurity,
.inode_setsecurity = selinux_inode_setsecurity,
.inode_listsecurity = selinux_inode_listsecurity,
Implement file posix capabilities Implement file posix capabilities. This allows programs to be given a subset of root's powers regardless of who runs them, without having to use setuid and giving the binary all of root's powers. This version works with Kaigai Kohei's userspace tools, found at http://www.kaigai.gr.jp/index.php. For more information on how to use this patch, Chris Friedhoff has posted a nice page at http://www.friedhoff.org/fscaps.html. Changelog: Nov 27: Incorporate fixes from Andrew Morton (security-introduce-file-caps-tweaks and security-introduce-file-caps-warning-fix) Fix Kconfig dependency. Fix change signaling behavior when file caps are not compiled in. Nov 13: Integrate comments from Alexey: Remove CONFIG_ ifdef from capability.h, and use %zd for printing a size_t. Nov 13: Fix endianness warnings by sparse as suggested by Alexey Dobriyan. Nov 09: Address warnings of unused variables at cap_bprm_set_security when file capabilities are disabled, and simultaneously clean up the code a little, by pulling the new code into a helper function. Nov 08: For pointers to required userspace tools and how to use them, see http://www.friedhoff.org/fscaps.html. Nov 07: Fix the calculation of the highest bit checked in check_cap_sanity(). Nov 07: Allow file caps to be enabled without CONFIG_SECURITY, since capabilities are the default. Hook cap_task_setscheduler when !CONFIG_SECURITY. Move capable(TASK_KILL) to end of cap_task_kill to reduce audit messages. Nov 05: Add secondary calls in selinux/hooks.c to task_setioprio and task_setscheduler so that selinux and capabilities with file cap support can be stacked. Sep 05: As Seth Arnold points out, uid checks are out of place for capability code. Sep 01: Define task_setscheduler, task_setioprio, cap_task_kill, and task_setnice to make sure a user cannot affect a process in which they called a program with some fscaps. One remaining question is the note under task_setscheduler: are we ok with CAP_SYS_NICE being sufficient to confine a process to a cpuset? It is a semantic change, as without fsccaps, attach_task doesn't allow CAP_SYS_NICE to override the uid equivalence check. But since it uses security_task_setscheduler, which elsewhere is used where CAP_SYS_NICE can be used to override the uid equivalence check, fixing it might be tough. task_setscheduler note: this also controls cpuset:attach_task. Are we ok with CAP_SYS_NICE being used to confine to a cpuset? task_setioprio task_setnice sys_setpriority uses this (through set_one_prio) for another process. Need same checks as setrlimit Aug 21: Updated secureexec implementation to reflect the fact that euid and uid might be the same and nonzero, but the process might still have elevated caps. Aug 15: Handle endianness of xattrs. Enforce capability version match between kernel and disk. Enforce that no bits beyond the known max capability are set, else return -EPERM. With this extra processing, it may be worth reconsidering doing all the work at bprm_set_security rather than d_instantiate. Aug 10: Always call getxattr at bprm_set_security, rather than caching it at d_instantiate. [morgan@kernel.org: file-caps clean up for linux/capability.h] [bunk@kernel.org: unexport cap_inode_killpriv] Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Cc: Stephen Smalley <sds@tycho.nsa.gov> Cc: James Morris <jmorris@namei.org> Cc: Chris Wright <chrisw@sous-sol.org> Cc: Andrew Morgan <morgan@kernel.org> Signed-off-by: Andrew Morgan <morgan@kernel.org> Signed-off-by: Adrian Bunk <bunk@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 14:31:36 +08:00
.inode_need_killpriv = selinux_inode_need_killpriv,
.inode_killpriv = selinux_inode_killpriv,
.inode_getsecid = selinux_inode_getsecid,
.file_permission = selinux_file_permission,
.file_alloc_security = selinux_file_alloc_security,
.file_free_security = selinux_file_free_security,
.file_ioctl = selinux_file_ioctl,
.file_mmap = selinux_file_mmap,
.file_mprotect = selinux_file_mprotect,
.file_lock = selinux_file_lock,
.file_fcntl = selinux_file_fcntl,
.file_set_fowner = selinux_file_set_fowner,
.file_send_sigiotask = selinux_file_send_sigiotask,
.file_receive = selinux_file_receive,
.dentry_open = selinux_dentry_open,
.task_create = selinux_task_create,
.task_alloc_security = selinux_task_alloc_security,
.task_free_security = selinux_task_free_security,
.task_setuid = selinux_task_setuid,
.task_post_setuid = selinux_task_post_setuid,
.task_setgid = selinux_task_setgid,
.task_setpgid = selinux_task_setpgid,
.task_getpgid = selinux_task_getpgid,
.task_getsid = selinux_task_getsid,
.task_getsecid = selinux_task_getsecid,
.task_setgroups = selinux_task_setgroups,
.task_setnice = selinux_task_setnice,
.task_setioprio = selinux_task_setioprio,
.task_getioprio = selinux_task_getioprio,
.task_setrlimit = selinux_task_setrlimit,
.task_setscheduler = selinux_task_setscheduler,
.task_getscheduler = selinux_task_getscheduler,
.task_movememory = selinux_task_movememory,
.task_kill = selinux_task_kill,
.task_wait = selinux_task_wait,
.task_prctl = selinux_task_prctl,
.task_reparent_to_init = selinux_task_reparent_to_init,
.task_to_inode = selinux_task_to_inode,
.ipc_permission = selinux_ipc_permission,
.ipc_getsecid = selinux_ipc_getsecid,
.msg_msg_alloc_security = selinux_msg_msg_alloc_security,
.msg_msg_free_security = selinux_msg_msg_free_security,
.msg_queue_alloc_security = selinux_msg_queue_alloc_security,
.msg_queue_free_security = selinux_msg_queue_free_security,
.msg_queue_associate = selinux_msg_queue_associate,
.msg_queue_msgctl = selinux_msg_queue_msgctl,
.msg_queue_msgsnd = selinux_msg_queue_msgsnd,
.msg_queue_msgrcv = selinux_msg_queue_msgrcv,
.shm_alloc_security = selinux_shm_alloc_security,
.shm_free_security = selinux_shm_free_security,
.shm_associate = selinux_shm_associate,
.shm_shmctl = selinux_shm_shmctl,
.shm_shmat = selinux_shm_shmat,
.sem_alloc_security = selinux_sem_alloc_security,
.sem_free_security = selinux_sem_free_security,
.sem_associate = selinux_sem_associate,
.sem_semctl = selinux_sem_semctl,
.sem_semop = selinux_sem_semop,
.d_instantiate = selinux_d_instantiate,
.getprocattr = selinux_getprocattr,
.setprocattr = selinux_setprocattr,
.secid_to_secctx = selinux_secid_to_secctx,
.secctx_to_secid = selinux_secctx_to_secid,
.release_secctx = selinux_release_secctx,
.unix_stream_connect = selinux_socket_unix_stream_connect,
.unix_may_send = selinux_socket_unix_may_send,
.socket_create = selinux_socket_create,
.socket_post_create = selinux_socket_post_create,
.socket_bind = selinux_socket_bind,
.socket_connect = selinux_socket_connect,
.socket_listen = selinux_socket_listen,
.socket_accept = selinux_socket_accept,
.socket_sendmsg = selinux_socket_sendmsg,
.socket_recvmsg = selinux_socket_recvmsg,
.socket_getsockname = selinux_socket_getsockname,
.socket_getpeername = selinux_socket_getpeername,
.socket_getsockopt = selinux_socket_getsockopt,
.socket_setsockopt = selinux_socket_setsockopt,
.socket_shutdown = selinux_socket_shutdown,
.socket_sock_rcv_skb = selinux_socket_sock_rcv_skb,
[SECURITY]: TCP/UDP getpeersec This patch implements an application of the LSM-IPSec networking controls whereby an application can determine the label of the security association its TCP or UDP sockets are currently connected to via getsockopt and the auxiliary data mechanism of recvmsg. Patch purpose: This patch enables a security-aware application to retrieve the security context of an IPSec security association a particular TCP or UDP socket is using. The application can then use this security context to determine the security context for processing on behalf of the peer at the other end of this connection. In the case of UDP, the security context is for each individual packet. An example application is the inetd daemon, which could be modified to start daemons running at security contexts dependent on the remote client. Patch design approach: - Design for TCP The patch enables the SELinux LSM to set the peer security context for a socket based on the security context of the IPSec security association. The application may retrieve this context using getsockopt. When called, the kernel determines if the socket is a connected (TCP_ESTABLISHED) TCP socket and, if so, uses the dst_entry cache on the socket to retrieve the security associations. If a security association has a security context, the context string is returned, as for UNIX domain sockets. - Design for UDP Unlike TCP, UDP is connectionless. This requires a somewhat different API to retrieve the peer security context. With TCP, the peer security context stays the same throughout the connection, thus it can be retrieved at any time between when the connection is established and when it is torn down. With UDP, each read/write can have different peer and thus the security context might change every time. As a result the security context retrieval must be done TOGETHER with the packet retrieval. The solution is to build upon the existing Unix domain socket API for retrieving user credentials. Linux offers the API for obtaining user credentials via ancillary messages (i.e., out of band/control messages that are bundled together with a normal message). Patch implementation details: - Implementation for TCP The security context can be retrieved by applications using getsockopt with the existing SO_PEERSEC flag. As an example (ignoring error checking): getsockopt(sockfd, SOL_SOCKET, SO_PEERSEC, optbuf, &optlen); printf("Socket peer context is: %s\n", optbuf); The SELinux function, selinux_socket_getpeersec, is extended to check for labeled security associations for connected (TCP_ESTABLISHED == sk->sk_state) TCP sockets only. If so, the socket has a dst_cache of struct dst_entry values that may refer to security associations. If these have security associations with security contexts, the security context is returned. getsockopt returns a buffer that contains a security context string or the buffer is unmodified. - Implementation for UDP To retrieve the security context, the application first indicates to the kernel such desire by setting the IP_PASSSEC option via getsockopt. Then the application retrieves the security context using the auxiliary data mechanism. An example server application for UDP should look like this: toggle = 1; toggle_len = sizeof(toggle); setsockopt(sockfd, SOL_IP, IP_PASSSEC, &toggle, &toggle_len); recvmsg(sockfd, &msg_hdr, 0); if (msg_hdr.msg_controllen > sizeof(struct cmsghdr)) { cmsg_hdr = CMSG_FIRSTHDR(&msg_hdr); if (cmsg_hdr->cmsg_len <= CMSG_LEN(sizeof(scontext)) && cmsg_hdr->cmsg_level == SOL_IP && cmsg_hdr->cmsg_type == SCM_SECURITY) { memcpy(&scontext, CMSG_DATA(cmsg_hdr), sizeof(scontext)); } } ip_setsockopt is enhanced with a new socket option IP_PASSSEC to allow a server socket to receive security context of the peer. A new ancillary message type SCM_SECURITY. When the packet is received we get the security context from the sec_path pointer which is contained in the sk_buff, and copy it to the ancillary message space. An additional LSM hook, selinux_socket_getpeersec_udp, is defined to retrieve the security context from the SELinux space. The existing function, selinux_socket_getpeersec does not suit our purpose, because the security context is copied directly to user space, rather than to kernel space. Testing: We have tested the patch by setting up TCP and UDP connections between applications on two machines using the IPSec policies that result in labeled security associations being built. For TCP, we can then extract the peer security context using getsockopt on either end. For UDP, the receiving end can retrieve the security context using the auxiliary data mechanism of recvmsg. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Acked-by: James Morris <jmorris@namei.org> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-03-21 14:41:23 +08:00
.socket_getpeersec_stream = selinux_socket_getpeersec_stream,
.socket_getpeersec_dgram = selinux_socket_getpeersec_dgram,
.sk_alloc_security = selinux_sk_alloc_security,
.sk_free_security = selinux_sk_free_security,
.sk_clone_security = selinux_sk_clone_security,
.sk_getsecid = selinux_sk_getsecid,
.sock_graft = selinux_sock_graft,
.inet_conn_request = selinux_inet_conn_request,
.inet_csk_clone = selinux_inet_csk_clone,
.inet_conn_established = selinux_inet_conn_established,
.req_classify_flow = selinux_req_classify_flow,
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
#ifdef CONFIG_SECURITY_NETWORK_XFRM
.xfrm_policy_alloc_security = selinux_xfrm_policy_alloc,
.xfrm_policy_clone_security = selinux_xfrm_policy_clone,
.xfrm_policy_free_security = selinux_xfrm_policy_free,
[LSM-IPsec]: SELinux Authorize This patch contains a fix for the previous patch that adds security contexts to IPsec policies and security associations. In the previous patch, no authorization (besides the check for write permissions to SAD and SPD) is required to delete IPsec policies and security assocations with security contexts. Thus a user authorized to change SAD and SPD can bypass the IPsec policy authorization by simply deleteing policies with security contexts. To fix this security hole, an additional authorization check is added for removing security policies and security associations with security contexts. Note that if no security context is supplied on add or present on policy to be deleted, the SELinux module allows the change unconditionally. The hook is called on deletion when no context is present, which we may want to change. At present, I left it up to the module. LSM changes: The patch adds two new LSM hooks: xfrm_policy_delete and xfrm_state_delete. The new hooks are necessary to authorize deletion of IPsec policies that have security contexts. The existing hooks xfrm_policy_free and xfrm_state_free lack the context to do the authorization, so I decided to split authorization of deletion and memory management of security data, as is typical in the LSM interface. Use: The new delete hooks are checked when xfrm_policy or xfrm_state are deleted by either the xfrm_user interface (xfrm_get_policy, xfrm_del_sa) or the pfkey interface (pfkey_spddelete, pfkey_delete). SELinux changes: The new policy_delete and state_delete functions are added. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Acked-by: James Morris <jmorris@namei.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-09 14:39:49 +08:00
.xfrm_policy_delete_security = selinux_xfrm_policy_delete,
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
.xfrm_state_alloc_security = selinux_xfrm_state_alloc,
.xfrm_state_free_security = selinux_xfrm_state_free,
[LSM-IPsec]: SELinux Authorize This patch contains a fix for the previous patch that adds security contexts to IPsec policies and security associations. In the previous patch, no authorization (besides the check for write permissions to SAD and SPD) is required to delete IPsec policies and security assocations with security contexts. Thus a user authorized to change SAD and SPD can bypass the IPsec policy authorization by simply deleteing policies with security contexts. To fix this security hole, an additional authorization check is added for removing security policies and security associations with security contexts. Note that if no security context is supplied on add or present on policy to be deleted, the SELinux module allows the change unconditionally. The hook is called on deletion when no context is present, which we may want to change. At present, I left it up to the module. LSM changes: The patch adds two new LSM hooks: xfrm_policy_delete and xfrm_state_delete. The new hooks are necessary to authorize deletion of IPsec policies that have security contexts. The existing hooks xfrm_policy_free and xfrm_state_free lack the context to do the authorization, so I decided to split authorization of deletion and memory management of security data, as is typical in the LSM interface. Use: The new delete hooks are checked when xfrm_policy or xfrm_state are deleted by either the xfrm_user interface (xfrm_get_policy, xfrm_del_sa) or the pfkey interface (pfkey_spddelete, pfkey_delete). SELinux changes: The new policy_delete and state_delete functions are added. Signed-off-by: Catherine Zhang <cxzhang@watson.ibm.com> Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Acked-by: James Morris <jmorris@namei.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-09 14:39:49 +08:00
.xfrm_state_delete_security = selinux_xfrm_state_delete,
.xfrm_policy_lookup = selinux_xfrm_policy_lookup,
.xfrm_state_pol_flow_match = selinux_xfrm_state_pol_flow_match,
.xfrm_decode_session = selinux_xfrm_decode_session,
#endif
#ifdef CONFIG_KEYS
.key_alloc = selinux_key_alloc,
.key_free = selinux_key_free,
.key_permission = selinux_key_permission,
.key_getsecurity = selinux_key_getsecurity,
#endif
#ifdef CONFIG_AUDIT
.audit_rule_init = selinux_audit_rule_init,
.audit_rule_known = selinux_audit_rule_known,
.audit_rule_match = selinux_audit_rule_match,
.audit_rule_free = selinux_audit_rule_free,
#endif
};
static __init int selinux_init(void)
{
struct task_security_struct *tsec;
if (!security_module_enable(&selinux_ops)) {
selinux_enabled = 0;
return 0;
}
if (!selinux_enabled) {
printk(KERN_INFO "SELinux: Disabled at boot.\n");
return 0;
}
printk(KERN_INFO "SELinux: Initializing.\n");
/* Set the security state for the initial task. */
if (task_alloc_security(current))
panic("SELinux: Failed to initialize initial task.\n");
tsec = current->security;
tsec->osid = tsec->sid = SECINITSID_KERNEL;
sel_inode_cache = kmem_cache_create("selinux_inode_security",
sizeof(struct inode_security_struct),
0, SLAB_PANIC, NULL);
avc_init();
secondary_ops = security_ops;
if (!secondary_ops)
panic("SELinux: No initial security operations\n");
if (register_security(&selinux_ops))
panic("SELinux: Unable to register with kernel.\n");
if (selinux_enforcing)
printk(KERN_DEBUG "SELinux: Starting in enforcing mode\n");
else
printk(KERN_DEBUG "SELinux: Starting in permissive mode\n");
return 0;
}
void selinux_complete_init(void)
{
printk(KERN_DEBUG "SELinux: Completing initialization.\n");
/* Set up any superblocks initialized prior to the policy load. */
printk(KERN_DEBUG "SELinux: Setting up existing superblocks.\n");
spin_lock(&sb_lock);
spin_lock(&sb_security_lock);
next_sb:
if (!list_empty(&superblock_security_head)) {
struct superblock_security_struct *sbsec =
list_entry(superblock_security_head.next,
struct superblock_security_struct,
list);
struct super_block *sb = sbsec->sb;
sb->s_count++;
spin_unlock(&sb_security_lock);
spin_unlock(&sb_lock);
down_read(&sb->s_umount);
if (sb->s_root)
superblock_doinit(sb, NULL);
drop_super(sb);
spin_lock(&sb_lock);
spin_lock(&sb_security_lock);
list_del_init(&sbsec->list);
goto next_sb;
}
spin_unlock(&sb_security_lock);
spin_unlock(&sb_lock);
}
/* SELinux requires early initialization in order to label
all processes and objects when they are created. */
security_initcall(selinux_init);
#if defined(CONFIG_NETFILTER)
static struct nf_hook_ops selinux_ipv4_ops[] = {
{
.hook = selinux_ipv4_postroute,
.owner = THIS_MODULE,
.pf = PF_INET,
.hooknum = NF_INET_POST_ROUTING,
.priority = NF_IP_PRI_SELINUX_LAST,
},
{
.hook = selinux_ipv4_forward,
.owner = THIS_MODULE,
.pf = PF_INET,
.hooknum = NF_INET_FORWARD,
.priority = NF_IP_PRI_SELINUX_FIRST,
},
{
.hook = selinux_ipv4_output,
.owner = THIS_MODULE,
.pf = PF_INET,
.hooknum = NF_INET_LOCAL_OUT,
.priority = NF_IP_PRI_SELINUX_FIRST,
}
};
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
static struct nf_hook_ops selinux_ipv6_ops[] = {
{
.hook = selinux_ipv6_postroute,
.owner = THIS_MODULE,
.pf = PF_INET6,
.hooknum = NF_INET_POST_ROUTING,
.priority = NF_IP6_PRI_SELINUX_LAST,
},
{
.hook = selinux_ipv6_forward,
.owner = THIS_MODULE,
.pf = PF_INET6,
.hooknum = NF_INET_FORWARD,
.priority = NF_IP6_PRI_SELINUX_FIRST,
}
};
#endif /* IPV6 */
static int __init selinux_nf_ip_init(void)
{
int err = 0;
if (!selinux_enabled)
goto out;
printk(KERN_DEBUG "SELinux: Registering netfilter hooks\n");
err = nf_register_hooks(selinux_ipv4_ops, ARRAY_SIZE(selinux_ipv4_ops));
if (err)
panic("SELinux: nf_register_hooks for IPv4: error %d\n", err);
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
err = nf_register_hooks(selinux_ipv6_ops, ARRAY_SIZE(selinux_ipv6_ops));
if (err)
panic("SELinux: nf_register_hooks for IPv6: error %d\n", err);
#endif /* IPV6 */
[LSM-IPSec]: Per-packet access control. This patch series implements per packet access control via the extension of the Linux Security Modules (LSM) interface by hooks in the XFRM and pfkey subsystems that leverage IPSec security associations to label packets. Extensions to the SELinux LSM are included that leverage the patch for this purpose. This patch implements the changes necessary to the SELinux LSM to create, deallocate, and use security contexts for policies (xfrm_policy) and security associations (xfrm_state) that enable control of a socket's ability to send and receive packets. Patch purpose: The patch is designed to enable the SELinux LSM to implement access control on individual packets based on the strongly authenticated IPSec security association. Such access controls augment the existing ones in SELinux based on network interface and IP address. The former are very coarse-grained, and the latter can be spoofed. By using IPSec, the SELinux can control access to remote hosts based on cryptographic keys generated using the IPSec mechanism. This enables access control on a per-machine basis or per-application if the remote machine is running the same mechanism and trusted to enforce the access control policy. Patch design approach: The patch's main function is to authorize a socket's access to a IPSec policy based on their security contexts. Since the communication is implemented by a security association, the patch ensures that the security association's negotiated and used have the same security context. The patch enables allocation and deallocation of such security contexts for policies and security associations. It also enables copying of the security context when policies are cloned. Lastly, the patch ensures that packets that are sent without using a IPSec security assocation with a security context are allowed to be sent in that manner. A presentation available at www.selinux-symposium.org/2005/presentations/session2/2-3-jaeger.pdf from the SELinux symposium describes the overall approach. Patch implementation details: The function which authorizes a socket to perform a requested operation (send/receive) on a IPSec policy (xfrm_policy) is selinux_xfrm_policy_lookup. The Netfilter and rcv_skb hooks ensure that if a IPSec SA with a securit y association has not been used, then the socket is allowed to send or receive the packet, respectively. The patch implements SELinux function for allocating security contexts when policies (xfrm_policy) are created via the pfkey or xfrm_user interfaces via selinux_xfrm_policy_alloc. When a security association is built, SELinux allocates the security context designated by the XFRM subsystem which is based on that of the authorized policy via selinux_xfrm_state_alloc. When a xfrm_policy is cloned, the security context of that policy, if any, is copied to the clone via selinux_xfrm_policy_clone. When a xfrm_policy or xfrm_state is freed, its security context, if any is also freed at selinux_xfrm_policy_free or selinux_xfrm_state_free. Testing: The SELinux authorization function is tested using ipsec-tools. We created policies and security associations with particular security contexts and added SELinux access control policy entries to verify the authorization decision. We also made sure that packets for which no security context was supplied (which either did or did not use security associations) were authorized using an unlabelled context. Signed-off-by: Trent Jaeger <tjaeger@cse.psu.edu> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2005-12-14 15:12:40 +08:00
out:
return err;
}
__initcall(selinux_nf_ip_init);
#ifdef CONFIG_SECURITY_SELINUX_DISABLE
static void selinux_nf_ip_exit(void)
{
printk(KERN_DEBUG "SELinux: Unregistering netfilter hooks\n");
nf_unregister_hooks(selinux_ipv4_ops, ARRAY_SIZE(selinux_ipv4_ops));
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
nf_unregister_hooks(selinux_ipv6_ops, ARRAY_SIZE(selinux_ipv6_ops));
#endif /* IPV6 */
}
#endif
#else /* CONFIG_NETFILTER */
#ifdef CONFIG_SECURITY_SELINUX_DISABLE
#define selinux_nf_ip_exit()
#endif
#endif /* CONFIG_NETFILTER */
#ifdef CONFIG_SECURITY_SELINUX_DISABLE
static int selinux_disabled;
int selinux_disable(void)
{
extern void exit_sel_fs(void);
if (ss_initialized) {
/* Not permitted after initial policy load. */
return -EINVAL;
}
if (selinux_disabled) {
/* Only do this once. */
return -EINVAL;
}
printk(KERN_INFO "SELinux: Disabled at runtime.\n");
selinux_disabled = 1;
selinux_enabled = 0;
/* Reset security_ops to the secondary module, dummy or capability. */
security_ops = secondary_ops;
/* Unregister netfilter hooks. */
selinux_nf_ip_exit();
/* Unregister selinuxfs. */
exit_sel_fs();
return 0;
}
#endif